JP6834544B2 - Sensor abnormality diagnostic device - Google Patents

Description

The present invention relates to a sensor abnormality diagnostic device.

Conventionally, as seen in Patent Document 1, an abnormality diagnosis device of two current sensors for detecting a current value flowing through a rotary electric machine is known. Each of the two current sensors detects the phase current value at the same detection location of the rotary electric machine. The abnormality diagnosis device acquires the phase current value detected at the same timing by the two current sensors as the current detection value. The abnormality diagnostic device is based on a comparison between the approximate straight line drawn by the current detection values acquired in the coordinate system whose coordinate axes are the current detection values of the two acquired current sensors and the reference straight line, and the abnormality of the two current sensors. Diagnose the presence or absence. When the current amplitude of the phase current is equal to or less than a predetermined value, the abnormality diagnosis device prohibits the abnormality diagnosis of the current sensor in order to suppress a decrease in the abnormality diagnosis accuracy.

Japanese Unexamined Patent Publication No. 2011-151986

Noise may be mixed in the current detection value of the current sensor. When noise is mixed in, the approximate straight line drawn by the acquired current detection value may deviate from the reference straight line even when the current sensor is not abnormal. Here, when the phase current value at the same detection location fluctuates momentarily, the current amplitude becomes large, so that the abnormality diagnosis device allows the abnormality diagnosis of the current sensor. In such a situation where the abnormality diagnosis is permitted, if noise is mixed in the current detection value, it is erroneously diagnosed that the current sensor has an abnormality even though the current sensor has not occurred. There is a concern.

As described above, if the system is equipped with a current sensor that detects the current value flowing through a predetermined detection target, not limited to the current sensor that detects the phase current value of the rotating electric machine, it is said that the sensor has an abnormality. There is a concern that it will be misdiagnosed. Further, not only the current sensor but also a system including a sensor that detects a predetermined state amount of a detection target may be erroneously diagnosed as having an abnormality in the sensor as described above.

An object of the present invention is to provide a sensor abnormality diagnosis device capable of suppressing the occurrence of a situation in which an abnormality is erroneously diagnosed as having an abnormality in the sensor even though the abnormality has not occurred in the sensor.

The present invention is applied to a system including a target sensor, which is a sensor that detects a state quantity of a predetermined detection target, and the first acquisition unit that acquires the detection value of the target sensor as a sensor detection value, and the target sensor With the second acquisition unit that acquires the detection value of another sensor that is another sensor and detects the state amount of the same detection target as the detection target of the target sensor, or the correlation value of the detection value of the other sensor as a determination value. In a coordinate system having each of the sensor detection value and the judgment value as a coordinate axis, the target sensor is based on a comparison between a judgment line drawn by a plurality of sets of the sensor detection value and the judgment value and a reference line. A diagnosis unit for diagnosing an abnormality, an index value calculation unit for calculating a variation index value which is a value indicating the magnitude of variation of at least one of the sensor detection value and the determination value used in the diagnosis unit, and the variation. A prohibition unit for prohibiting the diagnosis of the target sensor by the diagnosis unit when the index value is equal to or less than the threshold value is provided.

In the present invention, the detection value of another sensor that detects the state quantity of the detection target that is different from the target sensor and is the same as the detection target of the target sensor, or the correlation value of the detection value of the other sensor is defined as the determination value. Has been done. The diagnostic unit of the present invention is drawn by a plurality of sets of sensor detection values and judgment values in a coordinate system having each of the sensor detection value acquired by the first acquisition unit and the judgment value acquired by the second acquisition unit as coordinate axes. Based on the comparison between the judgment line and the reference line, the abnormality of the target sensor is diagnosed.

Here, when the variation of the plurality of sensor detection values used for calculating the determination line is equal to or smaller than the variation of the noise included in the sensor detection value, the noise included in the sensor detection value is based on the determination line. It will deviate greatly from the line. Further, when the variation of a plurality of determination values used for calculating the determination line is equal to or smaller than the variation of the noise included in the determination value, the noise included in the determination value greatly deviates from the reference line. Will be made to. As a result, even though the target sensor is not abnormal, it can be erroneously diagnosed that the abnormality has occurred. On the other hand, when the variation of the plurality of sensor detection values used for calculating the judgment line is larger than the variation of the noise included in the sensor detection value, the noise included in the sensor detection value deviates the judgment line from the reference line. The degree of attempt is small. Further, when the variation of a plurality of determination values used for calculating the determination line is larger than the variation of the noise included in the determination value, the degree to which the noise included in the determination value tends to deviate the determination line from the reference line is high. small.

In view of this point, the index value calculation unit of the present invention calculates a variation index value which is a value indicating the magnitude of variation of at least one of the sensor detection value and the determination value used in the diagnosis unit. Then, the prohibition unit prohibits the diagnosis of the target sensor by the diagnosis unit because there is a high risk of erroneous diagnosis when the calculated variation index value is equal to or less than the threshold value. As a result, it is possible to suppress the occurrence of a situation in which the target sensor is erroneously diagnosed as having an abnormality even though the target sensor has no abnormality.

The overall block diagram of the motor control system which concerns on 1st Embodiment. The block diagram which shows the output voltage control processing of a boost converter. The figure for demonstrating the time ratio. The flowchart which shows the procedure of the abnormality diagnosis processing of a current sensor. The figure which shows the 2nd sampling data group. The figure which shows the 1st sampling data group. The figure which shows the sampling data group of a 2D coordinate system, and the reference straight line Lthα. The figure for demonstrating that the abnormality diagnosis accuracy is lowered in the comparative technique. The figure which shows the judgment straight line Ljdg when the standard deviation of a sensor noise and the standard deviation of the whole sampling data are equivalent. The figure which shows the judgment straight line Ljdg when the standard deviation of the whole sampling data is sufficiently large with respect to the standard deviation of a sensor noise. The overall block diagram of the motor control system which concerns on 2nd Embodiment. The flowchart which shows the procedure of the abnormality diagnosis processing of a current sensor. The figure which shows the 2nd sampling data group. The figure which shows the 1st sampling data group. The figure which shows the sampling data group of a 2D coordinate system, and the reference straight line Lthα. The overall block diagram of the motor control system which concerns on 4th Embodiment. The overall block diagram of the motor control system which concerns on 5th Embodiment. The overall block diagram of the motor control system which concerns on 6th Embodiment. The overall block diagram of the motor control system which concerns on 7th Embodiment. The overall block diagram of the motor control system which concerns on 8th Embodiment. The flowchart which shows the procedure of the abnormality diagnosis processing of the current sensor which concerns on 9th Embodiment. The block diagram which shows the reactor current estimation processing which concerns on 11th Embodiment. The block diagram which shows the reactor current estimation processing which concerns on 12th Embodiment. The block diagram which shows the reactor current estimation processing which concerns on 13th Embodiment. The flowchart which shows the correction process of the time ratio which concerns on 14th Embodiment.

(First Embodiment)
Hereinafter, a first embodiment embodying the abnormality diagnostic apparatus according to the present invention will be described with reference to the drawings. In the present embodiment, the abnormality diagnosis device is mounted on a vehicle such as an electric vehicle or a hybrid vehicle provided with a rotating electric machine as an in-vehicle main engine.

As shown in FIG. 1, the vehicle-mounted control system includes a battery 10 as a DC power source, a boost converter 20, an inverter 30, a motor generator 40, and a control device 50. The battery 10 is a rechargeable power storage device. In the present embodiment, the motor generator 40 is an in-vehicle main engine, and its rotor is capable of transmitting power to drive wheels (not shown). As the motor generator 40, for example, a synchronous machine having a permanent magnet in the rotor can be used. The battery 10 and the boost converter 20 constitute a power supply system.

The boost converter 20 includes a reactor 21, a smoothing capacitor 22, an upper arm transformer switch Scp, and a lower arm transformer switch Scn. The boost converter 20 has a function of boosting the output voltage of the battery 10 up to a predetermined voltage. In this embodiment, each transformer switch Scp, Scn is a voltage-controlled semiconductor switching element, specifically an IGBT. Freewheel diodes Dcp and Dcn are connected in antiparallel to each transformer switch Scp and Scn.

A positive electrode bus Lp is connected to a collector which is a terminal on the high potential side of the upper arm transformer switch SCP. The collector of the lower arm transformer switch Scn is connected to the emitter which is the low potential side terminal of the upper arm transformer switch SCP. A negative electrode bus Ln is connected to the emitter of the lower arm transformer switch Scn. Each bus Lp, Ln is composed of, for example, a bus bar.

A smoothing capacitor 22 is connected in parallel to the series connection body of the upper arm transformer switch Scp and the lower arm transformer switch SCP. The first end of the reactor 21 is connected to the connection point between the upper arm transformer switch Scp and the lower arm transformer switch SCP. A positive electrode terminal of the battery 10 is connected to the second end of the reactor 21. The emitter of the lower arm transformer switch Scn is connected to the negative electrode terminal of the battery 10.

The input side of the inverter 30 is connected to the positive electrode bus Lp and the negative electrode bus Ln. The inverter 30 includes a series connection body of the upper arm switches Sup, Svp, Swp and the lower arm switches Sun, Svn, Swn for three phases. In the present embodiment, each switch Sup, Sun, Sbp, Svn, Swp, Swn is a voltage-controlled semiconductor switching element, and more specifically, an IGBT. Freewheel diodes Dup, Dun, Dbp, Dvn, Dwp, and Dwn are connected in antiparallel to each switch Sup, Sun, Spp, Svn, Swp, and Swn.

A positive electrode bus Lp is connected to a collector which is a terminal on the high potential side of each of the upper arm switches Sup, Svp, and Swp. A negative electrode bus Ln is connected to an emitter which is a low potential side terminal of each of the lower arm switches Sun, Svn, and Swn.

The first end of the U-phase winding 40U of the motor generator 40 is connected to the connection points of the U-phase upper and lower arm switches Sup and Sun. The first end of the V-phase winding 40V of the motor generator 40 is connected to the connection points of the V-phase upper and lower arm switches Svp and Svn. The first end of the W-phase winding 40W of the motor generator 40 is connected to the connection points of the W-phase upper and lower arm switches Swp and Swn. The second ends of each phase winding 40U, 40V, 40W are connected at the neutral point. The U, V, W phase windings 40U, 40V, 40W are offset by 120 ° from each other in terms of electrical angle.

The control system includes an electrical load 41. The positive electrode terminal of the electric load 41 is connected to an electric path connecting the positive electrode terminal of the battery 10 and the second end of the reactor 21. The negative electrode terminal of the electric load 41 is connected to an electric path connecting the negative electrode terminal of the battery 10 and the emitter of the lower arm transformer switch Scn. That is, the electric load 41 is connected in parallel to the battery 10. The electric load 41 includes, for example, at least one of a DCDC converter and an electric compressor. The DCDC converter steps down the output voltage of the battery 10 and supplies the stepped down voltage to another battery having an output voltage lower than that of the battery 10. The electric compressor circulates the refrigerant of the refrigeration cycle that constitutes the in-vehicle air conditioner.

The control system includes a first reactor current sensor 60a, a second reactor current sensor 60b, an input voltage sensor 61, an output voltage sensor 62, a phase current sensor 63, a rotation position sensor 64, and a battery current sensor 65. The first reactor current sensor 60a detects the current value flowing through the reactor 21 as the first reactor current detection value ILr1. The second reactor current sensor 60b detects the current value flowing through the reactor 21 as the second reactor current detection value ILr2. The input voltage sensor 61 detects the output voltage of the battery 10 as the input voltage detection value Vin. The output voltage sensor 62 detects the voltage between the terminals of the smoothing capacitor 22 as the bus voltage detection value Vsys. The phase current sensor 63 detects the phase currents of at least two of the U, V, and W phases. The rotation position sensor 64 is, for example, a resolver, and detects the rotation position of the rotor of the motor generator 40. The battery current sensor 65 detects the current value flowing through the battery 10 as the battery current detection value IB.

In the present embodiment, the first reactor current sensor 60a corresponds to the "target sensor", and the second reactor current sensor 60b corresponds to the "separate sensor". The first reactor current sensor 60a detects the current value flowing through the reactor 21 as the state quantity of the reactor 21. The second reactor current sensor 60b detects the current value of the reactor 21, which is the same detection target as the detection target of the first reactor current sensor 60a.

The detected value of each sensor is input to the control device 50. The control device 50 is mainly composed of a microcomputer, and operates a boost converter 20 and an inverter 30 in order to control the control amount of the motor generator 40 to the command value thereof. In the present embodiment, the control amount is torque and the command value is command torque Tcmd.

The control device 50 turns on / off the transformer switches Scpp and Scn constituting the boost converter 20 in order to feedback-control the bus voltage detection value Vsys detected by the output voltage sensor 62 to the target voltage value Vtgt. In the present embodiment, the upper arm transformer switch Scp and the lower arm transformer switch SCP are alternately turned on with a dead time in between.

The control device 50 turns on / off the switches Sup, Sun, Svp, Svn, Swp, and Swn of the inverter 30. The upper arm switches Sup, Sbp, Swp and the corresponding lower arm switches Sun, Svn, Swn are alternately turned on with a dead time in between.

Subsequently, with reference to FIG. 2, a process related to the boost converter 20 among the processes performed by the control device 50 will be described. In this embodiment, among both ends of the reactor 21, the current value flowing in the direction from the positive electrode terminal side of the battery 10 toward the connection point side of each transformer switch Scp and Scn is defined as positive.

The voltage deviation calculation unit 51 calculates the voltage deviation ΔV as a value obtained by subtracting the bus voltage detection value Vsys from the target voltage value Vtgt.

The voltage FB control unit 52 sets a target current value ILtgt, which is a target value of the current value flowing through the reactor 21, as an operation amount for feedback-controlling the bus voltage detection value Vsys to the target voltage value Vtgt based on the voltage deviation ΔV. calculate. In the present embodiment, the feedback control used by the voltage FB control unit 52 is proportional integration control.

The current deviation calculation unit 53 calculates the current deviation ΔI as a value obtained by subtracting the deviation calculation current value, which is either the first reactor current detection value ILr1 or the second reactor current detection value ILr2, from the target current value ILtgt.

The current FB control unit 54 calculates the time ratio duty as an operation amount for feedback-controlling the deviation calculation current value to the target current value ILtgt based on the current deviation ΔI. In the present embodiment, the feedback control used by the current FB control unit 54 is proportional integration control. As shown in FIG. 3, the time ratio duty is the ratio of the on-operation time Ton to one switching cycle Tsw of the lower arm transformer switch Scn. The time ratio duty is also a boost ratio expressed as "Vsys / Vin". In FIG. 3, the dead time is set to 0.

The control device 50 alternately turns on the upper arm transformer switch Scp and the lower arm transformer switch Scn based on the calculated time ratio duty. Specifically, for example, the control device 50 generates operation signals for the upper and lower arm transformer switches Scp and Scn by PWM processing based on the time ratio Duty and the magnitude comparison of the carrier signals, and moves up based on the generated operation signals. The arm transformer switch Scp and the lower arm transformer switch Scn are alternately turned on.

Subsequently, the abnormality diagnosis process of the first reactor current sensor 60a will be described with reference to FIG. This process is repeatedly executed by the control device 50, for example, at predetermined control cycles.

In this series of processes, first, in step S10, the past N second reactor current detection values ILr2 are acquired as the second sampling data groups Y [n−N + 1] to Y [n] shown in FIG. In FIG. 5, the current control cycle is n. Incidentally, in the present embodiment, the process of step S10 corresponds to the "second acquisition unit" that acquires the second reactor current detection value ILr2 as the determination value.

In the following step S11, the second standard, which is the standard deviation of the second reactor current detection values ILr2 for N pieces, is based on the second sampling data groups Y [n−N + 1] to Y [n] and the following equation (eq1). Calculate the deviation σy. In the following equation (eq1), Yave indicates the average value of the second sampling data group Y [n−N + 1] to Y [n]. The second standard deviation σy is an index value indicating the magnitude of variation from the average value Yave of the second sampling data group Y [n−N + 1] to Y [n].

続くステップＳ１２では、過去のＮ個分の第１リアクトル電流検出値ＩＬｒ１を、図６に示す第１サンプリングデータ群Ｘ［ｎ−Ｎ＋１］〜Ｘ［ｎ］として取得する。ちなみに本実施形態において、ステップＳ１２の処理が、センサ検出値としての第１リアクトル電流検出値ＩＬｒ１を取得する「第１取得部」に相当する。

In the following step S12, the first reactor current detection values ILr1 for the past N pieces are acquired as the first sampling data groups X [n−N + 1] to X [n] shown in FIG. Incidentally, in the present embodiment, the process of step S12 corresponds to the "first acquisition unit" that acquires the first reactor current detection value ILr1 as the sensor detection value.

It should be noted that the sampling timing of each sampling data X [na] of the first sampling data group and the sampling timing of each sampling data Y [na] of the corresponding second sampling data group must be synchronized. desirable.

In the following step S13, the first standard is the standard deviation of the first reactor current detection values ILr1 for N pieces based on the first sampling data groups X [n−N + 1] to X [n] and the following equation (eq2). Calculate the deviation σx. In the following equation (eq2), Xave represents the average value of the first sampling data group X [n−N + 1] to X [n]. The first standard deviation σx is an index value indicating the magnitude of variation from the average value Xave of the first sampling data groups X [n−N + 1] to X [n].

ちなみに本実施形態において、ステップＳ１１，Ｓ１３の処理が「指標値算出部」に相当する。また本実施形態において、第１，第２標準偏差σｘ，σｙが「第１，第２ばらつき指標値」に相当する。

Incidentally, in the present embodiment, the processes of steps S11 and S13 correspond to the "index value calculation unit". Further, in the present embodiment, the first and second standard deviations σx and σy correspond to the “first and second variation index values”.

In the following step S14, the logical sum (OR) of the first condition that the first standard deviation σx is larger than the first threshold value Isx and the second condition that the second standard deviation σy is larger than the second threshold value Isy. Determines if is true. This process is a process for determining whether or not to prohibit the abnormality diagnosis of the first reactor current sensor 60a. The first threshold value Itx may be set to the same value as the second threshold value Ithy, or may be set to a value different from the second threshold value Ithy.

If an affirmative determination is made in step S14, the process proceeds to step S15, the abnormality diagnosis is permitted, and the abnormality diagnosis process is performed. In this embodiment, the abnormality diagnosis method described in Patent Document 1 is used. Hereinafter, this abnormality diagnosis method will be described.

In step S15, in a two-dimensional coordinate system (see FIG. 7) having the first reactor current detection value ILr1 and the second reactor current detection value ILr2 as coordinate axes, N sets of first sampling data X [n-a] and The slope value SL of the determination straight line Ljdg, which is a simple regression line drawn by the second sampling data Y [na], is calculated. Here, the slope value SL is a simple regression line “B = k × A + cost” in which the first sampling data X [na] is the explanatory variable A and the second sampling data Y [na] is the objective variable B. It may be calculated as the coefficient k of. The simple regression line may be calculated by, for example, the well-known least squares method.

In step S15, it is diagnosed whether or not an abnormality has occurred in the first reactor current sensor 60a by comparing the calculated inclination value SL with the inclination value of the reference straight line Lthα shown in FIG. 7. In the present embodiment, the detection target of the first reactor current sensor 60a to be diagnosed is the same as the detection target of the second reactor current sensor 60b. Therefore, the slope value of the reference straight line Lthα is 1. Further, in the present embodiment, the reference straight line Lthα passes through the origin 0 of the coordinate system because it is assumed that the intercept is 0. Specifically, in step S15, it is determined whether or not “1-ΔS ≦ SL ≦ 1 + ΔS (ΔS> 0)” is satisfied. When it is determined that "1-ΔS ≦ SL ≦ 1 + ΔS" is satisfied, it is diagnosed that the first reactor current sensor 60a is normal. On the other hand, when it is determined that "1-ΔS ≦ SL ≦ 1 + ΔS" is not satisfied, it is diagnosed that an abnormality has occurred in the first reactor current sensor 60a.

If it is determined in step S14 that neither of the two conditions is satisfied, the process proceeds to step S16, and the abnormality diagnosis of the first reactor current sensor 60a is prohibited. The reasons for prohibiting abnormal diagnosis will be described below.

8 (b) and 8 (d) show that when the first and second reactor current sensors 60a and 60b are normal, noise is not mixed in the first reactor current detection value ILr1 and the second reactor current detection value ILr2. The judgment straight line Ljdg calculated for each of the cases of mixing is shown.

In FIG. 8A, when noise is not mixed in the first reactor current detection value ILr1 and the second reactor current detection value ILr2, the current detection values ILr1 and ILr2 undergo a momentary fluctuation in which the peak value is Ipeak. Show the case. As shown in FIG. 8A, when the current detection values ILr1 and ILr2 fluctuate instantaneously when noise is not mixed, the first reactor current detection value ILr1 and the second reactor current detection value ILr2 in the predetermined period Tjdg The determination straight line Ljdg calculated from the set of is as shown in FIG. 8 (b).

On the other hand, FIG. 8C shows a momentary peak value of Ipeak in each of the current detection values ILr1 and ILr2 when noise is mixed in the first reactor current detection value ILr1 and the second reactor current detection value ILr2. The case where various fluctuations occur is shown. When noise is mixed as shown in FIG. 8 (c), the determination straight line Ljdg calculated from the set of the first reactor current detection value ILr1 and the second reactor current detection value ILr2 in the predetermined period Tjdg is shown in FIG. As shown in, the deviation from the reference straight line Lthα is large. Therefore, when noise is mixed in, there is a concern that an abnormality may be erroneously diagnosed even though the first reactor current sensor 60a is normal.

Here, as shown in FIG. 9, the standard deviation of each current detection value ILr1 and ILr2 used for calculating the determination straight line Ljdg is equal to or the standard deviation of the noise included in each current detection value ILr1 and ILr2. If it is smaller than, the determination straight line Ljdg is affected by noise and greatly deviates from the reference straight line Lthα.

On the other hand, as shown in FIG. 10, the standard deviation of each current detection value ILr1 and ILr2 used for calculating the determination straight line Ljdg is the standard deviation of the noise included in each current detection value ILr1 and ILr2, and each current is detected. When the value is sufficiently larger than the standard deviation of the noise mixed in the values ILr1 and ILr2, the influence of the noise on the determination straight line Ljdg is small. Therefore, the determination straight line Ljdg substantially coincides with the reference straight line Lthα.

Focusing on this point, the process of step S14 of FIG. 4 is provided. FIG. 10 shows a situation in which either the first condition that the first standard deviation σx is larger than the first threshold value Isx and the second condition that the second standard deviation σy is larger than the second threshold value Isy are satisfied. It is highly probable that the situation shown will occur. Therefore, in the present embodiment, the abnormality diagnosis is permitted when a positive judgment is made in step S14, and the abnormality diagnosis is prohibited when a negative judgment is made in step S14.

By the way, in the present embodiment, the configuration in which the abnormality diagnosis is permitted by satisfying either the first condition or the second condition is the case where an abnormality occurs in either the first reactor current sensor 60a or the second reactor current sensor 60b. Even so, it is intended to allow abnormal diagnosis.

The first threshold Isthx is set based on the standard deviation of the noise mixed in the first reactor current detection value ILr1. The first threshold value Isx may be set to a value larger than the standard deviation of the noise mixed in the first reactor current detection value ILr1. Specifically, for example, the first threshold value Isx is twice the standard deviation of the noise mixed in the first reactor current detection value ILr1, or twice the standard deviation of the noise mixed in the first reactor current detection value ILr1. It may be set to a value larger than. The noise standard deviation used to set the first threshold Isx may be, for example, the noise standard deviation measured in advance when designing the control device, or the first reactor acquired online each time. It may be the standard deviation of noise calculated based on the current detection value ILr1.

The second threshold value Ity is set based on the standard deviation of the noise mixed in the second reactor current detection value ILr2. The second threshold value It may be set to a value larger than the standard deviation of the noise mixed in the second reactor current detection value ILr2. Specifically, for example, the second threshold value Ithy is twice the standard deviation of the noise mixed in the second reactor current detection value ILr2, or twice the standard deviation of the noise mixed in the second reactor current detection value ILr2. It may be set to a value larger than. The noise standard deviation used to set the second threshold Isy may be, for example, the noise standard deviation measured in advance when designing the control device, or the second reactor acquired online each time. It may be the standard deviation of noise calculated based on the current detection value ILr2.

As described above, in the present embodiment, when a negative determination is made in step S14 of FIG. 4, the abnormality diagnosis of the first reactor current sensor 60a is prohibited. As a result, it is possible to suppress the occurrence of a situation in which the first reactor current sensor 60a is erroneously diagnosed as having an abnormality even though the abnormality has not occurred.

(Second Embodiment)
Hereinafter, the second embodiment will be described with reference to the drawings, focusing on the differences from the first embodiment. In the present embodiment, as shown in FIG. 11, the control system includes one reactor current sensor 66 as a target sensor for detecting the current value flowing through the reactor 21. Note that, in FIG. 11, the same components as those shown in FIG. 1 above are designated by the same reference numerals for convenience.

The reactor current sensor 66 detects the current value flowing through the reactor 21 as the reactor current detection value ILr. The detected value of the reactor current sensor 66 is input to the control device 50.

Subsequently, the abnormality diagnosis process of the reactor current sensor 66 according to the present embodiment will be described with reference to FIG. In the present embodiment, the control device 50 performs an abnormality diagnosis process based on the reactor current detection value ILr and the reactor current estimated value ILest which is an estimated value of the current value flowing through the reactor 21. The process shown in FIG. 12 is repeatedly executed by the control device 50, for example, at predetermined control cycles.

In this series of processing, first, in step S20, the estimated reactor current values Irest for N past N pieces are calculated, and the calculated reactor current estimated values IREST are shown in FIG. 13 in the second sampling data group Y [n−N + 1] to Obtained as Y [n]. Incidentally, in the present embodiment, the process of step S20 corresponds to the "second acquisition unit" for acquiring the reactor current estimated value Irest as the determination value.

In the present embodiment, the reactor current estimated value Irest is calculated based on the motor power estimated value Pmot, the input voltage detection value Vin, which is the estimated value of the power of the motor generator 40, and the following equation (eq3).

なお、モータ電力推定値Ｐｍｏｔは、例えば、下式（ｅｑ４）に示されるように、モータジェネレータ４０のトルク推定値Ｔｅと、モータジェネレータ４０の回転速度Ｎｒとに基づいて算出されればよい。ここで、トルク推定値Ｔｅは、例えば、相電流センサ６３及び回転位置センサ６４の検出値に基づいて算出されればよい。

続くステップＳ２１では、第２サンプリングデータ群Ｙ［ｎ−Ｎ＋１］〜Ｙ［ｎ］及び上式（ｅｑ１）に基づいて、Ｎ個分のリアクトル電流推定値ＩＬｅｓｔの標準偏差である第２標準偏差σｙを算出する。

The motor power estimated value Pmot may be calculated based on, for example, the torque estimated value Te of the motor generator 40 and the rotation speed Nr of the motor generator 40, as shown in the following equation (eq4). Here, the torque estimation value Te may be calculated based on, for example, the detection values of the phase current sensor 63 and the rotation position sensor 64.

In the following step S21, the second standard deviation σy, which is the standard deviation of N reactor current estimates ILest, based on the second sampling data groups Y [n−N + 1] to Y [n] and the above equation (eq1). Is calculated.

In the following step S22, the past N reactor current detection values ILr are acquired as the first sampling data groups X [n−N + 1] to X [n] shown in FIG. Incidentally, in the present embodiment, the process of step S22 corresponds to the "first acquisition unit" for acquiring the reactor current detection value ILr as the sensor detection value.

In the following step S23, the first standard deviation σx, which is the standard deviation of N reactor current detection values ILr, is based on the first sampling data groups X [n−N + 1] to X [n] and the above equation (eq2). Is calculated.

In the following step S24, the logical sum of the first condition that the first standard deviation σx is larger than the first threshold value Isx and the second condition that the second standard deviation σy is larger than the second threshold value Isy is true. Determine if it exists. This process is a process for determining whether or not to prohibit the abnormality diagnosis of the reactor current sensor 66.

By the way, in the present embodiment, the configuration in which the abnormality diagnosis is permitted when either the first condition or the second condition is satisfied is the calculation process of the reactor current estimated value ILEST due to the failure of the microcomputer or the like constituting the control device 50. This is to enable abnormality diagnosis even when the above cannot be performed properly. Even when the current value actually flowing through the reactor 21 changes due to a failure of the microcomputer or the like, a sticking abnormality in which the estimated current current value Irest does not change may occur.

The first threshold value Isx is set based on the standard deviation of noise mixed in the reactor current detection value ILr. The first threshold value Isx may be set to a value larger than the standard deviation of the noise mixed in the reactor current detection value ILr. The noise standard deviation used to set the first threshold Isx may be, for example, the noise standard deviation measured in advance when designing the control device, or the reactor current detection acquired each time online. It may be the standard deviation of noise calculated based on the value ILr.

The second threshold Ithy is set based on the standard deviation of the noise mixed in the reactor current estimated value IREST. The second threshold value It may be set to a value larger than the standard deviation of the noise mixed in the reactor current estimated value IREST. The noise standard deviation used to set the second threshold Isy may be, for example, the noise standard deviation measured in advance at the time of designing the control device, or is acquired online each time and the reactor current. It may be the standard deviation of noise calculated based on the detection value of each sensor for calculating the estimated value ILest.

If an affirmative determination is made in step S24, the process proceeds to step S25, the abnormality diagnosis is permitted, and the abnormality diagnosis process is performed. Also in this embodiment, the abnormality diagnosis method described in step S14 of FIG. 4 above is used.

Specifically, in step S25, in a two-dimensional coordinate system (see FIG. 15) having each of the reactor current detection value ILr and the reactor current estimation value ILest as coordinate axes, N sets of first sampling data X [na] and the first sampling data X [na]. 2 The slope value SL of the determination straight line Ljdg, which is a simple regression straight line drawn by the sampling data Y [na], is calculated. Then, by comparing the calculated inclination value SL with the inclination value of the reference straight line Lthα shown in FIG. 15, it is diagnosed whether or not an abnormality has occurred in the reactor current sensor 66. In the present embodiment, the reactor current estimated value Irest is an estimated current value flowing through the detection target of the reactor current sensor 66 to be diagnosed. Therefore, the slope value of the reference straight line Lthα is 1. When it is determined in step S25 that "1-ΔS ≦ SL ≦ 1 + ΔS" is satisfied, it is diagnosed that the reactor current sensor 66 is normal. On the other hand, when it is determined that "1-ΔS ≦ SL ≦ 1 + ΔS" is not satisfied, it is diagnosed that an abnormality has occurred in the reactor current sensor 66.

If it is determined in step S24 that neither of the two conditions is satisfied, the process proceeds to step S26, and the abnormality diagnosis of the reactor current sensor 66 is prohibited.

In the present embodiment described above, the abnormality diagnosis is performed using the reactor current estimated value Irest. Therefore, it is possible to perform abnormality diagnosis of the reactor current sensor 66 while reducing the number of current sensors that detect the current value flowing through the reactor 21 and reducing the cost of the control system.

(Third Embodiment)
Hereinafter, the third embodiment will be described with reference to the drawings, focusing on the differences from the second embodiment. In the present embodiment, the battery current detection value IB is used instead of the reactor current estimated value IREST for the abnormality diagnosis of the reactor current sensor 66. In this case, the battery current sensor 65 corresponds to the "correlation sensor", and the battery current detection value IB corresponds to the "determination value".

Specifically, in step S20 of FIG. 12, the past N battery current detection values IB are acquired as the second sampling data group Y [n−N + 1] to Y [n]. After that, in step S25, in a two-dimensional coordinate system having each of the reactor current detection value ILr and the battery current detection value IB as the coordinate axes, N sets of the first sampling data X [na] and the second sampling data Y [n]. -A] calculates the slope value SL of the determination straight line Ljdg, which is a simple regression line drawn by.

In the present embodiment, it is desirable that the control device 50 performs the abnormality diagnosis process on the premise that the battery current sensor 65 is normal. Further, when the current value flowing through the electric load 41 is defined as the load current value IC, the control device 50 changes the load current value IC when the electric load 41 operates and the load current value IC is larger than 0. It is desirable to acquire the past N battery current detection values IB as the second sampling data group Y [n−N + 1] to Y [n] on condition that is equal to or less than a predetermined fluctuation amount.

According to the present embodiment described above, the same effect as that of the second embodiment can be obtained.

(Fourth Embodiment)
Hereinafter, the fourth embodiment will be described with reference to the drawings, focusing on the differences from the second embodiment. In the present embodiment, the abnormality diagnosis target is not the reactor current sensor 66 but the phase current sensor. FIG. 16 illustrates the first-phase current sensor 67a and the second-phase current sensor 67b that detect the U-phase current among the phase current sensors. In the present embodiment, the first-phase current sensor 67a corresponds to the "target sensor", and the second-phase current sensor 67b corresponds to the "separate sensor". Note that, in FIG. 16, the same components as those shown in FIG. 11 above are designated by the same reference numerals for convenience.

In the present embodiment, the control device 50 performs the abnormality diagnosis process of the first phase current sensor 67a by the same method as the abnormality diagnosis method described in the first embodiment. In this process, the detected values of the past N first-phase current sensors 67a are acquired as the first sampling data groups X [n−N + 1] to X [n], and the past N second-phase currents are obtained. The detected value of the sensor 67b may be acquired as the second sampling data group Y [n−N + 1] to Y [n].

According to the present embodiment described above, the same effect as that of the first embodiment can be obtained.

(Fifth Embodiment)
Hereinafter, the fifth embodiment will be described with reference to the drawings, focusing on the differences from the first embodiment. In the present embodiment, as shown in FIG. 17, the configuration of the boost converter 20 is changed. In FIG. 17, the same reference numerals are given to the same configurations as those shown in FIG. 1 above for convenience.

The boost converter 20 includes a first reactor 21a, a second reactor 21b, a smoothing capacitor 22, a first upper arm transformer switch Scap, a first lower arm transformer switch Scan, a second upper arm transformer switch Scbp, and a second lower arm transformer switch Scbn. It has.
In the present embodiment, each transformer switch Scap, Scan, Scbp, Scbn is an IGBT. Freewheel diodes Dcap, Dcan, Dcbp, and Dcbn are connected in antiparallel to each of the transformer switches Scap, Scan, Scbp, and Scbn.

A positive electrode bus Lp is connected to the collectors of the first and second upper arm transformer switches Scap and Scbn. The collector of the first lower arm transformer switch Scan is connected to the emitter of the first upper arm transformer switch Scap, and the collector of the second lower arm transformer switch Scbn is connected to the emitter of the second upper arm transformer switch Scbp. There is. The negative electrode bus Ln is connected to the emitters of the first and second lower arm transformer switches Scan and Scbn.

The first end of the first reactor 21a is connected to the connection point between the first upper arm transformer switch Scap and the first lower arm transformer switch Scan, and the second upper arm transformer switch Scbp and the second lower arm transformer switch Scbn The first end of the second reactor 21b is connected to the connection point of. The positive electrode terminal of the battery 10 is connected to the second end of the first reactor 21a and the second reactor 21b. Emitters of the first and second lower arm transformer switches Scan and Scbn are connected to the negative electrode terminal of the battery 10.

The control system includes a first reactor current sensor 68a and a second reactor current sensor 68b. Each of the reactor current sensors 68a and 68b detects the current value flowing in the electric path connecting the connection points of the second end of the first reactor 21a and the second end of the second reactor 21b and the positive electrode terminal of the battery 10. Therefore, each of the reactor current sensors 68a and 68b detects the total value of the current values flowing through the first reactor 21a and the second reactor 21b as the reactor current detection value ILr. The detected values of the reactor current sensors 68a and 68b are input to the control device 50.

In the present embodiment, the control device 50 performs the abnormality diagnosis process of the first reactor current sensor 68a by the same method as the abnormality diagnosis method described in the first embodiment. The detected values of the first reactor current sensors 68a for the past N pieces are acquired as the first sampling data groups X [n−N + 1] to X [n], and the detection values of the second reactor current sensors 68b for the past N pieces are detected. The value may be acquired as the second sampling data group Y [n−N + 1] to Y [n]. Incidentally, in the present embodiment, the first reactor current sensor 68a corresponds to the "target sensor", and the second reactor current sensor 68b corresponds to the "separate sensor".

According to the present embodiment described above, the same effect as that of the first embodiment can be obtained.

Incidentally, in the present embodiment, the control system may not include the second reactor current sensor 68b. In this case, the control device 50 may diagnose the presence or absence of an abnormality in the first reactor current sensor 68a by the same abnormality diagnosis process as in the second embodiment.

(Sixth Embodiment)
Hereinafter, the sixth embodiment will be described with reference to the drawings, focusing on the differences from the fifth embodiment. In this embodiment, as shown in FIG. 18, the detection location by the current sensor is changed. Note that, in FIG. 18, the same components as those shown in FIG. 17 above are designated by the same reference numerals for convenience.

The control system includes a first reactor current sensor 69a and a second reactor current sensor 69b. The first reactor current sensor 69a detects the current value flowing through the first reactor 21a as the first reactor current detection value ILr1. The second reactor current sensor 69b detects the current value flowing through the second reactor 21b as the second reactor current detection value ILr2. The detected values of the first reactor current sensor 69a and the second reactor current sensor 69b are input to the control device 50.

In the present embodiment, the current deviation calculation unit 53 of FIG. 2 described the current deviation ΔI as a value obtained by subtracting the added value of the first reactor current detection value ILr1 and the second reactor current detection value ILr2 from the target current value ILtgt. Is calculated.

In the present embodiment, it is assumed that the current value flowing through the first reactor 21a and the current value flowing through the second reactor 21b are equal to each other when the boost converter 20 is operating. More specifically, it is assumed that the DC component of the current value flowing through the first reactor 21a and the DC component of the current value flowing through the second reactor 21b are equal.

In the present embodiment, the control device 50 may perform the abnormality diagnosis processing of the first reactor current sensor 69a and the second reactor current sensor 69b in the same manner as the abnormality diagnosis method described in the second embodiment. it can.

Specifically, when the control device 50 performs an abnormality diagnosis of the first reactor current sensor 69a, the first sampling data group X corresponding to the "sensor detection value" is set to the first reactor current detection value ILr1 for the past N pieces. Obtained as [n−N + 1] to X [n]. Further, the control device 50 calculates the second reactor current estimated value Irest2 for N past pieces, and uses the calculated second reactor current estimated value Irest2 for N pieces as the second sampling data group corresponding to the "determination value". Obtained as Y [n−N + 1] to Y [n]. Here, the second reactor current estimated value IREST2 is an estimated value of the current value flowing through the second reactor 21b, and may be calculated based on, for example, the following equation (eq5) or the following equation (eq6). The second reactor current estimated value ILest2 is an estimated value of the state quantity having a correlation with the first reactor current detected value ILr1.

制御装置５０は、第２リアクトル電流センサ６９ｂの異常診断を行う場合、過去のＮ個分の第２リアクトル電流検出値ＩＬｒ２を、「センサ検出値」に相当する第１サンプリングデータ群Ｘ［ｎ−Ｎ＋１］〜Ｘ［ｎ］として取得する。また制御装置５０は、過去のＮ個分の第１リアクトル電流推定値ＩＬｅｓｔ１を算出し、算出したＮ個分の第１リアクトル電流推定値ＩＬｅｓｔ１を、「判定値」に相当する第２サンプリングデータ群Ｙ［ｎ−Ｎ＋１］〜Ｙ［ｎ］として取得する。ここで、第１リアクトル電流推定値ＩＬｅｓｔ１は、第１リアクトル２１ａに流れる電流値の推定値であり、例えば、上式（ｅｑ５）又は上式（ｅｑ６）に基づいて算出されればよい。

When the control device 50 performs an abnormality diagnosis of the second reactor current sensor 69b, the control device 50 uses the past N second reactor current detection values ILr2 as the first sampling data group X [n-] corresponding to the "sensor detection value". Obtained as N + 1] to X [n]. Further, the control device 50 calculates the first reactor current estimated value Irest1 for N past pieces, and uses the calculated first reactor current estimated value Irest1 for N pieces as the second sampling data group corresponding to the "determination value". Obtained as Y [n−N + 1] to Y [n]. Here, the first reactor current estimated value IREST1 is an estimated value of the current value flowing through the first reactor 21a, and may be calculated based on, for example, the above equation (eq5) or the above equation (eq6).

According to the present embodiment described above, the same effect as that of the second embodiment can be obtained.

In the present embodiment, the control system includes a plurality of current sensors (for example, two) for detecting the current value flowing in the first reactor 21a, and a plurality of current sensors (for example, 2) for detecting the current value flowing in the second reactor 21b. One) You may have it. In this case, the control device 50 can perform an abnormality diagnosis of the current sensor corresponding to each of the reactors 21a and 21b by using the same abnormality diagnosis method as the abnormality diagnosis method of the first embodiment.

Further, in the present embodiment, the boost converter provided with two sets of upper and lower arm transformer switches has been described as an example, but the boost converter may be provided with three or more sets of upper and lower arm transformer switches. In this case, a reactor is individually provided corresponding to each of the upper and lower arm transformer switch sets, and a current sensor is individually provided corresponding to each reactor.

(7th Embodiment)
Hereinafter, the seventh embodiment will be described with reference to the drawings, focusing on the differences from the fourth embodiment. In this embodiment, as shown in FIG. 19, the control system includes a first battery 10a and a second battery 10b. Note that, in FIG. 19, the same components as those shown in FIG. 16 above are designated by the same reference numerals for convenience.

As shown in FIG. 19, the second end of the reactor 21 is connected to the positive electrode terminals of the first battery 10a and the second battery 10b, respectively. The emitter of the lower arm transformer switch Scn is connected to the negative electrode terminals of each of the first battery 10a and the second battery 10b.

The control system includes a first battery current sensor 70a and a second battery current sensor 70b. The first battery current sensor 70a detects the current value flowing through the first battery 10a as the first battery current detection value IB1, and the second battery current sensor 70b detects the current value flowing through the second battery 10b as the second battery current detection value IB1. Detect as value IB2. The detected values of the sensors 70a and 70b are input to the control device 50.

In the present embodiment, the control device 50 can perform the abnormality diagnosis process of the reactor current sensor 66 by the same method as the abnormality diagnosis method described in the second embodiment. In this case, the reactor current estimated value IREST may be calculated based on, for example, the above equation (eq3) or the following equation (eq7).

また本実施形態において、制御装置５０は、第１バッテリ電流検出値ＩＢ１又は第２バッテリ電流検出値ＩＢ２と、リアクトル電流検出値ＩＬｒとに基づいて、リアクトル電流センサ６６の異常診断処理を行うことができる。この場合、第１，第２バッテリ電流センサ７０ａ，７０ｂが「相関センサ」に相当し、第１，第２バッテリ電流検出値ＩＢ１，ＩＢ２が「判定値」に相当する。

Further, in the present embodiment, the control device 50 may perform an abnormality diagnosis process of the reactor current sensor 66 based on the first battery current detection value IB1 or the second battery current detection value IB2 and the reactor current detection value ILr. it can. In this case, the first and second battery current sensors 70a and 70b correspond to the "correlation sensor", and the first and second battery current detection values IB1 and IB2 correspond to the "determination value".

(8th Embodiment)
Hereinafter, the eighth embodiment will be described with reference to the drawings, focusing on the differences from the sixth and seventh embodiments. In the present embodiment, the control system includes a first battery 10a individually provided corresponding to the first reactor 21a and a second battery 10b individually provided corresponding to the second reactor 21b. .. In FIG. 20, the same components as those shown in FIGS. 18 and 19 are designated by the same reference numerals for convenience.

As shown in FIG. 20, the second end of the first reactor 21a is connected to the positive electrode terminal of the first battery 10a, and the second end of the second reactor 21b is connected to the positive electrode terminal of the second battery 10b. ing. Emitters of the first and second lower arm transformer switches Scan and Scbn are connected to the negative electrode terminals of the first battery 10a and the second battery 10b, respectively.

In the present embodiment, the control device 50 may perform abnormality diagnosis of each of the first reactor current sensor 69a and the second reactor current sensor 69b by using, for example, the abnormality diagnosis method described in the second and third embodiments. it can.

In this embodiment, one of the first battery 10a and the second battery 10b may be a capacitor.

(9th Embodiment)
Hereinafter, the ninth embodiment will be described with reference to the drawings, focusing on the differences from the first embodiment. In the present embodiment, the control device 50 performs the abnormality diagnosis process by excluding a part of the sampling data.

FIG. 21 shows a procedure for abnormality diagnosis processing of the first reactor current sensor 60a according to the present embodiment. This process is repeatedly executed by the control device 50, for example, at predetermined control cycles. In FIG. 21, the same step numbers are assigned to the same processes as those shown in FIG. 4 above for convenience.

In this series of processes, after the processes of steps S10 and S12 are completed, the exclusion process is performed in step S30. This process will be described below.

It is determined by the set of the first sampling data X [na] and the second sampling data Y [na] in the coordinate system having each of the first reactor current detection value ILr1 and the second reactor current detection value ILr2 as the coordinate axes. A point is defined as a judgment point. Further, in this coordinate system, a set of the first sampling data X [na] and the second sampling data Y [na] in which the distance between the determination point and the reference straight line Lthα is larger than the predetermined distance Dis is defined as the exclusion target. Has been done.

In step S30, the exclusion target is excluded from the plurality of sets of the first sampling data X [na] and the second sampling data Y [na] acquired in steps S10 and S12.

In the following step S31, the second sampling data group Y [n−N + 1] to Y [n] acquired in step S10 is based on data other than the second sampling data excluded in the process of step S30. Calculate the standard deviation σy.

In the following step S32, among the first sampling data groups X [n−N + 1] to X [n] acquired in step S12, the first is based on the data other than the first sampling data excluded in the process of step S30. Calculate the standard deviation σx. After that, in the following step S14, the first and second standard deviations σx and σy calculated in steps S31 and S32 are used.

In step S15, among the second sampling data groups Y [n−N + 1] to Y [n] acquired in step S10, data other than the second sampling data excluded in the process of step S30 and in step S12. The slope value SL of the determination straight line Ljdg is calculated based on the acquired first sampling data groups X [n−N + 1] to X [n] other than the first sampling data excluded in the process of step S30. To do.

According to the present embodiment described above, the second sampling data groups Y [n−N + 1] to Y [n] acquired in step S10 and the first sampling data groups X [n−N + 1] to acquired in step S12. From X [n], the data of a predetermined ratio whose value is on the upper side and the data of a predetermined ratio whose value is on the lower side are excluded. As a result, it is possible to prevent the sampling data mixed with large noise from being used for calculating the standard deviation. As a result, it is possible to improve the accuracy of determining whether or not the abnormality diagnosis is prohibited.

(10th Embodiment)
Hereinafter, the tenth embodiment will be described with reference to the drawings, focusing on the differences from the ninth embodiment. In this embodiment, the processing method of step S30 in FIG. 21 above is changed. Specifically, in step S30, among the first sampling data groups X [n−N + 1] to X [n] acquired in step S12, the first sampling data of a predetermined ratio (for example, 10%) of the upper side and the lower side is used. , The first exclusion process of excluding from the first sampling data group X [n−N + 1] to X [n] acquired in step S12 is performed. Specifically, for example, among the first sampling data groups X [n−N + 1] to X [n] acquired in step S12, the first sampling data having a predetermined ratio from the maximum value side and a predetermined ratio from the minimum value side can be obtained. It is excluded from the first sampling data groups X [n−N + 1] to X [n] acquired in step S12. Further, in step S30, among the second sampling data groups Y [n−N + 1] to Y [n] acquired in step S10, the second sampling data of the above-mentioned predetermined proportions on the upper side and the lower side are acquired in step S10. The second exclusion process of excluding from the second sampling data group Y [n−N + 1] to Y [n] is performed. Specifically, for example, among the second sampling data groups Y [n−N + 1] to Y [n] acquired in step S10, the second sampling data having a predetermined ratio from the maximum value side and a predetermined ratio from the minimum value side can be obtained. It is excluded from the second sampling data Y [n−N + 1] to Y [n] acquired in step S10. As a result, the second sampling data paired with the excluded first sampling data is excluded from the second sampling data groups Y [n−N + 1] to Y [n].

In the following step S31, the second sampling data group Y [n−N + 1] to Y [n] acquired in step S10 is based on data other than the second sampling data excluded in the process of step S30. Calculate the standard deviation σy. In the following step S32, among the first sampling data groups X [n−N + 1] to X [n] acquired in step S12, the first is based on the data other than the first sampling data excluded in the process of step S30. Calculate the standard deviation σx. After that, in the following step S14, the first and second standard deviations σx and σy calculated in steps S31 and S32 are used.

According to the present embodiment described above, the sampling data in which a large amount of noise is mixed is the standard deviation by a simple process such as removing a predetermined ratio of data on the upper side and the lower side from the acquired sampling data group. It can be suppressed from being used in the calculation. As a result, it is possible to suppress an increase in the processing load of the control device 50 due to the abnormality diagnosis.

By the way, in the first exclusion process, the first sampling data group X [n−N + 1] to X [n] of the first sampling data group X [n−N + 1] to X [n] acquired in step S12 from the first sampling data of a predetermined ratio of either the upper side or the lower side. It may be a process of excluding from. Further, in the second exclusion process, the second sampling data group Y [n−N + 1] to Y [n] obtained in step S10 for a predetermined ratio of the second sampling data on either the upper side or the lower side. It may be a process of excluding from. Further, the predetermined ratio on the upper side and the predetermined ratio on the lower side may be set to different ratios.

(11th Embodiment)
Hereinafter, the eleventh embodiment will be described with reference to the drawings, focusing on the differences from the second embodiment. In the present embodiment, in order to improve the calculation accuracy of the second standard deviation σy and the abnormality diagnosis accuracy of the reactor current sensor 66, the method of estimating the current value flowing through the reactor 21 is changed. The control device 50 calculates the reactor current estimated value Irest in consideration of the change in the accumulated charge amount of the smoothing capacitor 22. Hereinafter, the current estimation method will be described.

The value of the current flowing through the reactor 21 increases or decreases according to the change in the accumulated charge amount of the smoothing capacitor 22. The amount of stored charge in the smoothing capacitor 22 varies, for example, due to a temporary increase in the power consumption of the motor generator 40. Therefore, in order to improve the estimation accuracy of the current flowing through the reactor 21, it is necessary to consider the change in the accumulated charge amount of the smoothing capacitor 22. Here, the reactor current estimated value Irest in consideration of the change in the accumulated charge amount of the smoothing capacitor 22 is expressed by the following equation (eq8).

上式（ｅｑ８）において、Ｐｍはモータジェネレータ４０の電力を示し、Ｃは平滑コンデンサ２２の静電容量を示す。上式（ｅｑ８）の右辺の第２項には、母線電圧検出値Ｖｓｙｓの時間微分値が含まれている。時間微分値は、母線電圧検出値Ｖｓｙｓのノイズの影響を受けやすいため、時間微分値がリアクトル電流推定値ＩＬｅｓｔの算出に用いられると、電流推定精度が低下し得る。したがって、時間微分値がリアクトル電流推定値ＩＬｅｓｔの算出に用いられるのは好ましくない。そこで本実施形態では、微分演算が行われない電流推定方法が用いられている。

In the above equation (eq8), Pm indicates the power of the motor generator 40, and C indicates the capacitance of the smoothing capacitor 22. The second term on the right side of the above equation (eq8) includes the time derivative value of the bus voltage detection value Vsys. Since the time derivative is easily affected by the noise of the bus voltage detection value Vsys, if the time derivative is used to calculate the reactor current estimate ILest, the current estimation accuracy may decrease. Therefore, it is not preferable that the time derivative value is used to calculate the reactor current estimated value ILest. Therefore, in the present embodiment, a current estimation method in which no differential calculation is performed is used.

By modifying the above equation (eq8), the following equation (eq9) is derived.

上式（ｅｑ９）において、「Ｄｕｔｙ＝Ｖｉｎ／Ｖｓｙｓ」及び「Ｉｍ＝Ｐｍ／Ｖｓｙｓ」である。Ｉｍをモータ電流値と称すこととする。上式（ｅｑ９）の両辺を積分すると、下式（ｅｑ１０）が導かれる。

In the above equation (eq9), “Duty = Vin / Vsys” and “Im = Pm / Vsys”. Im is referred to as a motor current value. By integrating both sides of the above equation (eq9), the following equation (eq10) is derived.

上式（ｅｑ１０）において、Ｖｃａｌを、平滑コンデンサ２２の端子間電圧の推定値である母線電圧推定値と称すこととする。上式（ｅｑ１０）は、「Ｖｓｙｓ＝Ｖｃａｌ」となるＩＬｅｓｔがリアクトル電流推定値であることを示している。本実施形態では、母線電圧推定値Ｖｃａｌを算出するための積分時間の制約等の都合により、リアクトル電流推定値ＩＬｅｓｔを逐次算出する補償器が用いられる。補償器が備えられる構成を図２２に示す。図２２は、制御装置５０が行う処理のうち、電流推定に関する処理を示すブロック図である。

In the above equation (eq10), Vcal is referred to as a bus voltage estimated value which is an estimated value of the voltage between terminals of the smoothing capacitor 22. The above equation (eq10) indicates that the Irest such that “Vsys = Vcal” is the estimated reactor current. In the present embodiment, a compensator for sequentially calculating the reactor current estimated value IREST is used due to the limitation of the integration time for calculating the bus voltage estimated value Vcal. The configuration in which the compensator is provided is shown in FIG. FIG. 22 is a block diagram showing a process related to current estimation among the processes performed by the control device 50.

The motor current estimation unit 80 calculates the motor current estimation value Imest, which is the current estimation value supplied from the boost converter 20 to the inverter 30. In the present embodiment, the boost converter 20 calculates the motor current estimated value Imest based on the command torque Tcmd, the bus voltage detection value Vsys, and the rotation speed Nr of the motor generator 40, as in the following equation (eq11).

モデル推定部８１は、電流ＦＢ制御部５４により算出された時比率Ｄｕｔｙと、モータ電流推定部８０により算出されたモータ電流推定値Ｉｍｅｓｔと、補償器８３により算出されたリアクトル電流推定値ＩＬｅｓｔとを、下式（ｅｑ１２）に示すコンバータモデルに入力することにより、母線電圧推定値Ｖｃａｌを算出する。

The model estimation unit 81 has a time ratio duty calculated by the current FB control unit 54, a motor current estimated value Imest calculated by the motor current estimation unit 80, and a reactor current estimated value Irest calculated by the compensator 83. , The bus voltage estimated value Vcal is calculated by inputting to the converter model shown in the following equation (eq12).

上式（ｅｑ１２）は、平滑コンデンサ２２の蓄積電荷量の変化を考慮した電流推定式である上式（ｅｑ８）に基づいて導かれたものである。このため、上式（ｅｑ１２）で表されるコンバータモデルは、母線電圧推定値Ｖｃａｌが平滑コンデンサ２２の蓄積電荷量の変化に依存するように構成されたモデルである。

The above equation (eq12) is derived based on the above equation (eq8), which is a current estimation equation considering a change in the accumulated charge amount of the smoothing capacitor 22. Therefore, the converter model represented by the above equation (eq12) is a model in which the bus voltage estimated value Vcal depends on the change in the accumulated charge amount of the smoothing capacitor 22.

The error calculation unit 82 calculates the voltage estimation error Verr as a value obtained by subtracting the bus voltage estimation value Vcal from the bus voltage detection value Vsys.

The compensator 83 calculates the reactor current estimated value Irest as an operation amount for feedback-controlling the voltage estimation error Verr to 0. In the present embodiment, the feedback control used in the compensator 83 is proportional integration control as shown in the following equation (eq13). The following equation (eq13) is the transfer function Co of the compensator 83. In the following equation (eq13), Kpa indicates the proportional gain, Kia indicates the integral gain, and s indicates the Laplace operator.

補償器８３により算出されたリアクトル電流推定値ＩＬｅｓｔは、モデル推定部８１に入力され、母線電圧推定値Ｖｃａｌの算出に用いられる。なお、リアクトル２１の両端のうち、バッテリ１０の正極端子側から各変圧スイッチＳｃｐ，Ｓｃｎの接続点側へと向かう方向が正となるようにリアクトル電流推定値ＩＬｅｓｔの極性が定義されている。

The reactor current estimated value Irest calculated by the compensator 83 is input to the model estimation unit 81 and used for calculating the bus voltage estimated value Vcal. Of both ends of the reactor 21, the polarity of the reactor current estimated value IREST is defined so that the direction from the positive electrode terminal side of the battery 10 toward the connection point side of each transformer switch Scp and Scn is positive.

According to the current estimation process described above, when “Vcal <Vsys”, the reactor current estimated value Irest is increased assuming that the reactor current estimated value IREST is smaller than the current value actually flowing through the reactor 21. On the other hand, when "Vcal> Vsys", the reactor current estimated value Irest is reduced, assuming that the reactor current estimated value IREST is larger than the current value actually flowing through the reactor 21. As a result, the reactor current estimated value IREST converges to its true value. When the reactor current estimated value Irest converges to its true value, “Vcal = Vsys”.

The processing of the motor current estimation unit 80, the processing of the model estimation unit 81, the processing of the error calculation unit 82, and the processing of the compensator 83 are sequentially executed in one control cycle. As a result, the reactor current estimated value IREST is sequentially calculated for each control cycle. The calculated reactor current estimate ILest is used in the process shown in FIG. 12 above.

Incidentally, the current estimation method shown in FIG. 22 may be used in the sixth embodiment. In this case, if the control device 50 sets 1/2 of the reactor current estimated value IREST 1 estimated by the process shown in FIG. 22 as the first reactor current estimated value Irest1 and the second reactor current estimated value IREST2, respectively. Good.

(12th Embodiment)
Hereinafter, the twelfth embodiment will be described with reference to the drawings, focusing on the differences from the eleventh embodiment. In the present embodiment, the motor current estimation unit 84 shown in FIG. 23 calculates the motor current estimated value Imest in consideration of the losses in each of the inverter 30 and the motor generator 40. Note that, in FIG. 23, the same components as those shown in FIG. 22 above are designated by the same reference numerals for convenience.

As shown in the following equation (eq14), the motor current estimation unit 84 uses the motor power estimation value Pmot, which is an estimated value of the power of the motor generator 40, the bus voltage detection value Vsys, and the power loss Plus in each of the inverter 30 and the motor generator 40. Based on this, the estimated motor current Imest is calculated. Here, the power loss Plus includes resistance loss and iron loss in the motor generator 40, and conduction loss and switching loss in the inverter 30.

以上説明した本実施形態によれば、リアクトル２１に流れる電流の推定精度をより高めることができる。

According to the present embodiment described above, the estimation accuracy of the current flowing through the reactor 21 can be further improved.

(13th Embodiment)
Hereinafter, the thirteenth embodiment will be described with reference to the drawings, focusing on the differences from the eleventh embodiment. In this embodiment, as shown in FIG. 24, the reactor current detection value ILr detected by the reactor current sensor 66 is used for estimating the current flowing through the reactor 21. In FIG. 24, the same components as those shown in FIG. 22 above are designated by the same reference numerals for convenience.

The compensator 85 calculates a current correction value Icor for correcting the reactor current detection value ILr as an operation amount for feedback-controlling the voltage estimation error Verr to 0. In the present embodiment, the feedback control used in the compensator 85 is proportional integration control as shown in the following equation (eq15). The following equation (eq15) is the transfer function Co of the compensator 85. In the following equation (eq15), Kpb indicates a proportional gain and Kib indicates an integrated gain.

補正部８６は、リアクトル電流検出値ＩＬｒに電流補正値Ｉｃｏｒを加算した値としてリアクトル電流推定値ＩＬｅｓｔを算出する。補正部８６から出力されたリアクトル電流推定値ＩＬｅｓｔは、モデル推定部８１に入力され、母線電圧推定値Ｖｃａｌの算出に用いられる。

The correction unit 86 calculates the reactor current estimated value ILest as a value obtained by adding the current correction value Icor to the reactor current detection value ILr. The reactor current estimated value Irest output from the correction unit 86 is input to the model estimation unit 81 and used for calculating the bus voltage estimated value Vcal.

The processing of the motor current estimation unit 80, the processing of the model estimation unit 81, the processing of the error calculation unit 82, the processing of the compensator 85, and the processing of the correction unit 86 are sequentially executed in one control cycle. As a result, the reactor current estimated value IREST is sequentially calculated for each control cycle.

According to the current estimation process described above, when the reactor current sensor 66 does not have an abnormality, even if the reactor current detection value ILr fluctuates, the followability of the reactor current estimation value Irest to the fluctuation can be improved. This is because, when the reactor current sensor 66 is not abnormal, the reactor current detection value ILr and the reactor current estimation value ILest substantially match, so that the model estimation unit 81, the compensator 85, and the like are not affected by the delay. Is.

(14th Embodiment)
Hereinafter, the 14th embodiment will be described with reference to the drawings, focusing on the differences from the 11th embodiment. In the present embodiment, the time ratio duty used in the model estimation unit 81 is corrected based on the dead time DT.

The actual time ratio changes by the dead time depending on the polarity of the current value flowing through the reactor 21. Therefore, the control device 50 corrects the time ratio duty calculated by the process shown in FIG. 2 above based on the dead time DT.

FIG. 25 shows the procedure of the time ratio correction process based on the dead time. This process is repeatedly executed by the control device 50, for example, at predetermined control cycles.

In step S40, it is determined whether or not the reactor current detection value ILr is 0 or more. This process is a process for determining the polarity of the current value flowing through the reactor 21. In the present embodiment, as described above, among both ends of the reactor 21, the current value flowing in the direction from the positive electrode terminal side of the battery 10 toward the connection point side of each transformer switch Scp and Scn is defined as positive.

If an affirmative judgment is made in step S40, the process proceeds to step S41, and the effective time ratio Dtyf is calculated by adding the dead time correction value dmod represented by the following equation (eq16) to the time ratio Duty. In the following equation (eq16), Tc indicates the period of the carrier signal used when generating the operation signal of each transformer switch Scp, Scn.

ステップＳ４０において否定判定した場合には、リアクトル２１に流れる電流値の極性が負であると判定し、ステップＳ４２に進む。ステップＳ４２では、時比率Ｄｕｔｙからデットタイム補正値ｄｍｏｄを減算した値として、実効時比率Ｄｔｙｆを算出する。

If a negative determination is made in step S40, it is determined that the polarity of the current value flowing through the reactor 21 is negative, and the process proceeds to step S42. In step S42, the effective time ratio Dtyf is calculated as a value obtained by subtracting the dead time correction value dmod from the time ratio Duty.

The effective time ratio Dtyf calculated in steps S41 and S42 is used in the model estimation unit 81. As a result, the estimation accuracy of the current value flowing through the reactor 21 can be further improved.

(15th Embodiment)
Hereinafter, the fifteenth embodiment will be described with reference to the drawings, focusing on the differences from the eleventh embodiment. In the present embodiment, the calculation method of the reactor current estimated value ILEST is changed. Specifically, the control device 50 uses the following equation (eq17) based on the above equation (eq8), the motor power estimated value Pmot, the input voltage detection value Vin, the bus voltage detection value Vsys, and the time differential value of the bus voltage detection value " By inputting "dVsys / dt", the reactor current estimated value IREST is calculated.

以上説明した本実施形態によれば、上記第１１実施形態の効果に準じた効果を得ることができる。

According to the present embodiment described above, it is possible to obtain an effect similar to the effect of the eleventh embodiment.

(Other embodiments)
In addition, each of the above-described embodiments may be modified as follows.

In the twelfth embodiment, the second term on the right side of the above equation (eq14) may be deleted to calculate the motor current estimated value Imest in which the power loss loss is ignored.

In the eleventh embodiment, the motor current estimated value Imest is applied to the voltage command vector Vdq of the dq coordinate system applied from the inverter 30 to the motor generator 40 and the motor generator 40 as expressed by the following equation (eq18). It may be calculated based on the inner product of the current command vector Idq of the flowing dq coordinate system. In this case, for example, the motor current estimated value Imest can be calculated even in a control device in which the torque estimation process of the motor generator 40 is not implemented.

・制御システムにモータジェネレータが複数備えられていてもよい。例えば、制御システムにモータジェネレータが２つ備えられる場合、上記第１１実施形態におけるモータ電流推定値Ｉｍｅｓｔは、例えば下式（ｅｑ１９）に基づいて算出されてもよい。なお下式（ｅｑ１９）において、Ｔｅ１，Ｔｅ２は第１，第２モータジェネレータのトルク推定値を示し、Ｎｒ１，Ｎｒ２は第１，第２モータジェネレータの回転速度を示す。

-The control system may be equipped with a plurality of motor generators. For example, when the control system is provided with two motor generators, the motor current estimated value Imest in the eleventh embodiment may be calculated based on, for example, the following equation (eq19). In the following equation (eq19), Te1 and Te2 indicate the torque estimates of the first and second motor generators, and Nr1 and Nr2 indicate the rotation speeds of the first and second motor generators.

・上記第１１〜第１４実施形態において、補償器で用いられるフィードバック制御としては、比例積分制御に限らず、例えば、積分制御又は比例制御であってもよい。また、比例積分制御、積分制御及び比例制御に微分制御がさらに含まれていてもよい。

-In the above 11th to 14th embodiments, the feedback control used in the compensator is not limited to the proportional integral control, and may be, for example, integral control or proportional control. Further, the proportional integral control, the integral control and the proportional control may further include the differential control.

In the process of step S14 of FIG. 4 and the process of step S24 of FIG. 12, the first condition that the first standard deviation σx is larger than the first threshold value Isx, and the second standard deviation σy is larger than the second threshold value Isy. It may be a process of determining whether or not the logical product (AND) of the second condition of being large is true.

In step S14 of FIG. 4 and step S24 of FIG. 12, either the first condition or the second condition may be eliminated.

The reference line is not limited to the reference straight line Lthα, and may be a reference curve in which the second sampling data Y [na] is uniquely determined with respect to the first sampling data X [na]. In this case, for example, in the first embodiment, the control device 50 compares the determination curve calculated based on the first sampling data group and the second sampling data group with the reference curve to obtain the first reactor current sensor. An abnormality diagnosis of 60a may be performed. For example, when the control device 50 determines that the area of the region sandwiched between the determination curve and the reference curve is larger than the predetermined area, the first reactor current sensor 60a is abnormal because the deviation between the determination curve and the reference curve is large. May be diagnosed as occurring.

-The variation index value is not limited to the standard deviation, but may be a variance. In this case, the calculation of the square root can be eliminated, and the calculation load of the control device 50 can be reduced.

Further, as the variation index value, for example, the tile index T shown in the following equation (eq20) may be used. The following equation (eq20) shows the tile index for the first sampling data.

また、ばらつき指標値として、例えば下式（ｅｑ２１）に示すエントロピー指数Ｉαが用いられてもよい。なお下式（ｅｑ２１）には、第１サンプリングデータについてのエントロピー指数を示す。

Further, as the variation index value, for example, the entropy index Iα represented by the following equation (eq21) may be used. The following equation (eq21) shows the entropy index for the first sampling data.

・診断対象となる対象センサとしては、電流センサに限らず、例えば電圧センサであってもよい。この場合、制御システムが、平滑コンデンサ２２の端子間電圧を検出する第１，第２出力電圧センサを備えているとする。この構成において、先の図４に示した診断方法を用いて第１出力電圧センサの異常の有無を診断できる。ここでは、第１出力電圧センサの複数の検出値が第１サンプリングデータ群として取得され、第２出力電圧センサの複数の検出値が第２サンプリングデータ群として取得されればよい。

-The target sensor to be diagnosed is not limited to the current sensor, but may be, for example, a voltage sensor. In this case, it is assumed that the control system includes first and second output voltage sensors that detect the voltage between the terminals of the smoothing capacitor 22. In this configuration, the presence or absence of abnormality in the first output voltage sensor can be diagnosed by using the diagnostic method shown in FIG. 4 above. Here, a plurality of detected values of the first output voltage sensor may be acquired as the first sampling data group, and a plurality of detected values of the second output voltage sensor may be acquired as the second sampling data group.

-The load device in which electric power is supplied from the battery via the boost converter is not limited to the inverter and the motor generator.

-The power source connected to the input side of the boost converter is not limited to the battery, but may be another power storage device such as a capacitor.

-The system to which the abnormality diagnosis device is applied is not limited to the control system of the rotary electric machine. Further, the system to which the abnormality diagnosis device is applied is not limited to the system mounted on the vehicle.

20 ... Boost converter, 21 ... Reactor, 50 ... Control device, 60a, 60b ... First and second reactor current sensors.

Claims (14)

It is applied to a system provided with a target sensor (60a; 66, etc.) which is a sensor for detecting the state quantity of a predetermined detection target (21; 21a, 21b).
The first acquisition unit (S12; S22) that acquires the detection value of the target sensor as the sensor detection value (ILr1; ILr), and
The detection value (ILr2, etc.) of another sensor (60b, etc.) that is different from the target sensor and detects the state quantity of the same detection target as the detection target of the target sensor, or the detection value of the other sensor. The second acquisition unit (S10; S20) that acquires the correlation value (Irest, etc.) as the determination value, and
In a coordinate system having each of the sensor detection value and the judgment value as a coordinate axis, based on a comparison between a judgment line (Ljdg) drawn by a plurality of sets of the sensor detection value and the judgment value and a reference line (Lthα). , The diagnostic unit (S15; S25) for diagnosing the abnormality of the target sensor, and
If the value indicating the magnitude of the variation and the variation index value, at least one of said distribution index value of the sensor detection value and the determination value used in the diagnostic unit (sigma] x, .sigma.y) index value calculating section for calculating a (S11, S13; S21, S23; S31, S32) and
The variation index value of noise that the variation index value calculated by the index value calculation unit is mixed with the value of the sensor detection value and the determination value to be calculated by the index value calculation unit. A sensor abnormality diagnosis device including a prohibition unit (S14, S16; S24, S26) that prohibits the diagnosis unit from diagnosing the target sensor when the threshold value (Itsx, Ithy) or less is set based on the above. The sensor abnormality diagnosis device according to claim 1, wherein the second acquisition unit (S10) acquires the detection value of the other sensor as the determination value. The system includes a correlation sensor (65, etc.) that is a sensor different from the target sensor and that detects a state quantity that has a correlation with the state quantity of the detection target.
The sensor abnormality diagnosis device according to claim 1, wherein the second acquisition unit acquires the detection value of the correlation sensor as the determination value. The second acquisition unit (S20) estimates the state quantity of the detection target that is the same as the detection target of the target sensor as the correlation value (Irest), and acquires the estimated correlation value as the determination value according to claim 1. The sensor abnormality diagnostic device described. The system includes a boost converter (20) that boosts and outputs a voltage input from a power supply (10).
The boost converter includes a reactor (21) that can be connected to the power supply and a smoothing capacitor (22) that is connected to the output side of the boost converter.
The target sensor (60a) is a current sensor that detects a current value flowing through the reactor.
The system includes a voltage sensor (62) that detects the voltage value of the smoothing capacitor.
The second acquisition unit is a current value flowing through the reactor based on a voltage detection value (Vsys) which is a voltage value detected by the voltage sensor, and depends on a change in the accumulated charge amount of the smoothing capacitor. The sensor abnormality diagnostic device according to claim 4, wherein the current value is estimated as the correlation value. The sensor abnormality diagnostic device according to claim 5, wherein the second acquisition unit estimates the correlation value based on the voltage detection value and the differential value of the voltage detection value. The second acquisition unit is a model in which the current value (Irest) flowing through the reactor and the output current value (Imest) of the boost converter are input, and the output voltage value (Vcal) of the boost converter is output. The abnormality diagnosis device for a sensor according to claim 6, wherein the correlation value is estimated based on a converter model configured such that the output voltage value depends on a change in the accumulated charge amount of the smoothing capacitor. The second acquisition unit calculates a voltage estimated value, which is an estimated value of the output voltage value, based on the converter model, and uses the calculated voltage estimated value as an operation amount for feedback control to the voltage detected value. The sensor abnormality diagnostic device according to claim 7, wherein the correlation value is estimated. The second acquisition unit
A voltage estimated value, which is an estimated value of the output voltage value, is calculated based on the converter model, and the calculated voltage estimated value is detected by the target sensor as an operation amount for feedback control to the voltage detected value. The correction value calculation unit (85) that calculates the correction value (Icor) that corrects the current detection value, which is the current value,
The abnormality diagnosis of the sensor according to claim 7, which includes a correction unit (86) that estimates the correlation value by correcting the current detection value based on the correction value calculated by the correction value calculation unit. apparatus. The second acquisition unit estimates the state quantity having a correlation with the state quantity of the detection target as the correlation value (Irest1 and IREST2), and acquires the estimated correlation value as the determination value. Sensor abnormality diagnostic device. The first acquisition unit acquires a plurality of the sensor detection values, and obtains the plurality of sensor detection values.
The second acquisition unit acquires a plurality of the determination values and obtains the plurality of determination values.
In the coordinate system, a point determined by the set of the sensor detection value and the determination value is defined as a determination point.
In the coordinate system, a set of the sensor detection value and the judgment value in which the distance between the judgment point and the reference line is larger than a predetermined distance is defined as an exclusion target.
An exclusion unit (S30) for excluding the exclusion target from a plurality of sets of the sensor detection value and the determination value acquired by the first acquisition unit and the second acquisition unit is provided.
The index value calculation unit (S31, S32) is based on the set of the sensor detection value and the determination value from which the exclusion target is excluded from the plurality of sets of the sensor detection value and the determination value, and the variation index. The sensor abnormality diagnostic device according to any one of claims 1 to 10, wherein the value is calculated. The first acquisition unit acquires a plurality of the sensor detection values, and obtains the plurality of sensor detection values.
The second acquisition unit acquires a plurality of the determination values and obtains the plurality of determination values.
Among the plurality of sensor detection values acquired by the first acquisition unit, the sensor detection value on at least one of the upper side and the lower side is set among the plurality of sensor detection values acquired by the first acquisition unit. Of the plurality of determination values acquired by the second acquisition unit, at least one of the upper side and the lower side of the determination values is the determination value of the plurality of determination values acquired by the second acquisition unit. Equipped with an exclusion unit that eliminates from the inside
The index value calculation unit calculates the variation index value based on a set other than the set of the sensor detection value and the determination value excluded by the exclusion unit among the plurality of sets of the sensor detection value and the determination value. The sensor abnormality diagnostic device according to any one of claims 1 to 10 for calculation. The index value calculation unit calculates the distribution index value before Symbol sensor detection value (sigma] x), the distribution index value of the judgment values and (.sigma.y),
Wherein a said threshold value corresponding to the distribution index value of the sensor detection value, a value which the set based on the distribution index value of the noise mixed in the sensor detection value is defined as a first threshold value (Ithx), wherein The threshold value corresponding to the variation index value of the determination value, which is set based on the variation index value of the noise mixed in the determination value, is defined as the second threshold value (Ithy).
The prohibition unit is described by the diagnostic unit when the variation index value of the sensor detection value is equal to or less than the first threshold value and the variation index value of the determination value is equal to or less than the second threshold value. The sensor abnormality diagnosis device according to any one of claims 1 to 12, which prohibits diagnosis of the target sensor. The sensor abnormality diagnostic device according to any one of claims 1 to 13, wherein the variation index value is a standard deviation or a variance.

Images