JP2022144063A - Sensor-to-sensor error angle estimation method and position identification method - Google Patents

Abstract

To provide a method for estimating an error angle between sensors from information derived by using values detected by sensors without depending on arithmetic processing inside a system.SOLUTION: A sensor-to-sensor error angle estimation method includes acquiring the states before and after attitude change by using a first sensor and a second sensor mounted for estimating the state of a system, and calculating an estimated error value between the calculated state of the system and a true value to estimate an error angle between the sensors. The method includes varying the error angle in a pseudo manner through internal arithmetic processing and approximating, by a function, the variation relationship of the calculated state of the system with the true value. The method includes, by using the approximated function, solving the function so that the error between the calculated value and the true value becomes 0 to estimate the sensor-to-sensor error angle.SELECTED DRAWING: Figure 4

Description

The present invention is applicable to systems equipped with multiple sensors.

Accurately measuring and calibrating the alignment between sensors is important in a system that performs sensor fusion. However, after calibration, it is conceivable that the alignment between the sensors will deviate by a certain error angle while the system is in operation. For example, in a ground position identification system, there is a method of estimating a desired matrix using the least squares method using sensor data.

An accuracy of better than 200m in positioning using a portable star tracker, Meysam Izadmehr, and Mehdi Khakian Ghomi, New Astronomy, Vol.74, 2020, doi:10.1016/j.newast.2019.04.004.

However, while the above approach is guaranteed to work in a particular system, the estimation method needs to be reconfigured if the system for which the sensor-to-sensor error angle is to be estimated changes. Accordingly, there is a need for a sensor-to-sensor error angle estimation method that can be applied even if the system's internal computational processes or the configuration itself changes.

In view of the above, the inter-sensor error angle estimation method according to the present invention includes:
A method of estimating a sensor-to-sensor error angle between a first sensor and a second sensor, comprising:
A housing equipped with a first sensor that acquires a first state of a mounted housing and a second sensor that acquires a second state different from the first state of the housing. a posture change step for changing the posture;
a state acquiring step of acquiring the first state and the second state by the first sensor and the second sensor before and after the posture changing step;
a state estimation step of estimating a state of the housing based on the first state and the second state obtained in the state obtaining step;
a true value acquisition step of acquiring a true value corresponding to the estimated value estimated in the state estimation step;
an estimated error deriving step of setting a pseudo error angle between the first sensor and the second sensor and deriving an estimated error value between the estimated value and the true value;
and an inter-sensor error angle estimation step of deriving the inter-sensor error angle using the derived estimated error value.

According to the present invention, a plausible inter-sensor error angle that minimizes the estimated value error is estimated by giving a pseudo error angle between sensors and deriving an estimated error value between the true value and the estimated value. This allows estimation of the sensor-to-sensor error angle independent of the internal computational processes or configuration of the system.

The figure which shows the structural example of the position identification system containing the error angle estimation method. FIG. 2 is a diagram for explaining a sensor coordinate system in the configuration example of FIG. 1; The figure which shows the relationship between a gravity direction vector and latitude longitude. 4 is a flowchart of an inter-sensor error angle estimation method in a position identification system; FIG. 5 is a flowchart showing an example of the procedure of step S3 in FIG. 4; FIG. FIG. 5 is a flowchart showing an example of the procedure of step S6 in FIG. 4; FIG. FIG. 5 is a flowchart showing an example of the procedure of step S7 in FIG. 4; FIG.

Preferred embodiments of the present invention are described below. It should be noted that the embodiments described below do not limit the content of the present invention described in the claims. Moreover, not all the configurations described below are essential constituent elements of the present invention. As an example of the method of estimating the error angle between sensors, a system for identifying the position of a spacecraft on a star with gravity such as a planet or a satellite is taken as an example.

1 Position Estimation Theory 1-1 Position Estimation System Configuration An example of the configuration of a system for identifying a position on a star in this embodiment is as shown in FIGS. 1 and 2. FIG. As shown in FIG. 1, the self-localization machine 200 includes an acceleration sensor 3 and a star tracker (STT) 5, and the calculation device 1 that processes the information obtained from each sensor is used to operate the self-localization machine 200. Estimate the planetary position of the machine. At this time, the position on the star is defined using latitude and longitude information.

As an example of the self-localization device 200 that constitutes the system of FIG. 1, there is a configuration in which the acceleration sensor 3 and the STT 5 are mounted on the casing 300 of the probe, as shown in FIG. A coordinate system is defined for each sensor, and the coordinate system fixed to the acceleration sensor (ACC) 3 is defined as an ACC coordinate system A, and the coordinate system fixed to the STT 5 is defined as an STT coordinate system. Also, in this configuration example, the STT coordinate system S and the body coordinate system are assumed to match. Even if they do not match, the results calculated in the SST coordinate system S can be converted into the body coordinate system.

As will be described later, a rotation matrix between the coordinate system S and the coordinate system A is required for position identification. However, various factors cause mechanical mounting errors in each sensor, which greatly affect the accuracy of position identification. Therefore, it is required to accurately estimate the rotation matrix between sensors. In this configuration example, the inter-sensor error angle is defined as an angle at which the ACC coordinate system A of the acceleration sensor 3 changes to the ACC coordinate system A' as shown in FIG. . Other details such as the configuration will be described later.

1-2 Method of Self-Position Identification from Gravity Vector A method of estimation using vectors will be described. This method calculates the gravitational acceleration vector g ^c =[g ^c _x, g ^c _y, g ^c _z ] with respect to the fixed coordinate system C of the star shown in FIG. calculate.

A gravitational acceleration vector is obtained from the output of the acceleration sensor 3 mounted on the self-localization machine 200 . At this time, the gravitational vector becomes the vector g ^A observed in the ACC coordinate system A of the acceleration sensor 3 . In order to estimate the self-position, the gravitational vector g ^A obtained by the acceleration sensor 3 is coordinate-transformed from the ACC coordinate system A to the fixed coordinate system C of the star, and the respective axial directions are converted to the equations (1) and (2). must be substituted with the magnitude of the gravitational vector of .

1-3 Derivation of rotation matrix from ACC coordinate system A to fixed star coordinate system C Rotation matrix R ^c/A from ACC coordinate system A to fixed star coordinate system C Using the rotation matrix R ^c/I to the system C and the rotation matrix R ^I/ A from the ACC coordinate system A to the inertial coordinate system I, it is represented by the following equation. Inertial coordinate system I can be defined as a coordinate system in which each axis is fixed in space, represented by a reference coordinate system such as the J2000 system if defined on the earth. is not limited to

Using the Greenwich sidereal time ξ, the rotation matrix R ^I/S from the STT coordinate system S to the inertial coordinate system I, and the rotation matrix R ^S/ A from the ACC coordinate system A to the STT coordinate system S, can be rewritten as

Here, the matrix _Rz is a coordinate rotation matrix about the z-axis, and the matrix _RNP is a rotation matrix that considers precessional nutation. Matrix _Rz (ξ) and matrix _RNP are calculated from the time information acquired from the clock 12 shown in FIG. Also, the matrix R ^I/S is a matrix obtained from the attitude information determined by the STT5. The matrix R ^S/A is a rotation matrix representing the relationship between the STT 5 and the acceleration sensor 3 measured in advance. The rotation matrix ^Rc/A can be obtained by using the rotation matrix obtained from the sensor information. Using the rotation matrix R ^{c /A} and the gravitational direction vector g ^A observed in the accelerometer coordinate system A, the gravitational direction vector g ^c with respect to the fixed coordinate system C of the star can be calculated as follows.

By substituting the elements of the gravitational direction vector of each axis obtained by the expression (5) into the expressions (1) and (2), the latitude and longitude on the star of the spacecraft equipped with the self-localization system 100 can be obtained. .

1-4 Consideration of inter-sensor error angle In order to perform position identification with high accuracy, the rotation matrix R ^S/A from the acceleration sensor coordinate system A to the STT coordinate system S in Equation (4) must be obtained with high accuracy before use. It is necessary to leave However, as described above, the rotation matrix may change due to various factors (for example, factors such as vibration). That is, the rotation matrix between the acceleration sensors 3 for the STT 5 changes from the matrix R ^S/A to R ^S/A' (FIG. 2).

When the self-localization system 100 is used on the earth, an angular error of approximately 0.01 degrees corresponds to a ground distance of approximately 1.0 km, which significantly reduces the accuracy of self-localization. Therefore, it is necessary to estimate the correction rotation matrix R ^A/A' when using it, and to calculate the gravitational direction vector for the fixed coordinate system C of the star by adding the correction rotation matrix R ^A/A' as shown in the following equation.

1-5 Inter-sensor Error Angle Estimation Method In this embodiment, as a method for estimating the inter-sensor error angle matrix ΔΘ=(Δθ _x , Δθ _y , Δθ _z ), a pseudo inter-sensor error angle matrix ΔΘ Substitute _p into the correction rotation matrix R ^A/A' , and record the matrix used for the function representing the relationship between the true value and the estimated error of the estimated value. For example, ^RA/A' can be represented by the rotation matrix below.

To the position identification method of this embodiment including the corrected rotation matrix R ^A/A' of equation (7), the gravity vector g ^A' measured by the acceleration sensor 3 having an installation error, and the posture information obtained from the STT 5 Substitute the matrix q and calculate the position on a star. At this time, the pseudo inter-sensor error angle (pseudo error angle) ΔΘ _p is changed by internal calculation, and the estimated error matrix ΔP between the acquired true position information (here, latitude φ _T and longitude λ _T ) is finally output to For acquisition of true position information (true value), for example, a relative VLBI method, which will be described later, is preferably used. Defining a series of these processes as a function f, it is represented by the following formula.

The relationship between this estimated error matrix ΔP and the values of the pseudo error angle matrix ΔΘ _p input to the function f is expressed by a function. From Equation (8), if the relationship between the estimated error matrix ΔP and the pseudo error angle ΔΘ _p can be represented by the following linear equation, the following equation is obtained.

By deriving a coefficient matrix G(q) _2×3 and a matrix H(q) _2×1 that satisfy the relationship of Equation (9), the inter-sensor error angle matrix ΔΘ is estimated. The coefficient matrix G(q) _2×3 and the matrix H(q) _2×1 are calculated for each orientation information matrix q by changing the orientation of the housing 300 on which the STT 5 and the acceleration sensor 3 are mounted. That is, the acceleration sensor 3 estimates the gravitational direction vector g ^A' with respect to the ACC coordinate system A', and the STT 5 uses the star vectors extracted from the captured image and the data of the star catalog stored in advance. The pseudo-error angle matrix ΔΘ _p is changed at each attitude while changing the attitude information matrix q, and the coefficient matrix G(q) _2×3 and the matrix H(q) _2×1 are calculated and recorded. Since the inter-sensor error angle matrix ΔΘ to be obtained is a value at which the estimated error matrix ΔP is 0, the solution is obtained by modifying Equation 9 and solving the following equation for the inter-sensor error angle ΔΘ.

In this case, the matrix G ⁻¹ _2n×3 is the pseudo-inverse of the matrix G _2n×3 .

2 Inter-sensor Error Angle Estimation Method 2-1 Flowchart of Inter-sensor Error Angle Estimation Method in Position Identification Method FIG. 4 is an example of a flowchart including an inter-sensor error angle estimation method in the position identification method according to this embodiment. An example of the procedure shown in FIG. 4 includes a method of estimating the error angle of each axis between the acceleration sensor 3 and the STT 5 mounted on the self-localization system 100 on a certain star.

It should be noted that this technique can also be implemented in a completely different system that is not intended for self-localization. For example, the system targeted by this method is a system that estimates a certain state using information from multiple sensors such as satellites and vehicles. If the attitude dependency is taken into consideration, it is sufficient to have means to change the attitude of the system.

As shown in FIG. 4, first, the orientation matrix q of the housing 300 on which the acceleration sensor 3 and the STT 5 are mounted is set to a preset initial value _q1 (step S1). Assume that the posture of the housing 300 can be changed by a mechanism such as a gimbal. After that, the initial value of the pseudo-error angular matrix ΔΘ _p is set to 0 degrees for each of the x, y, and z axes in step S2, and calculation by self-position identification (step S3) is performed. Detailed processing of self-localization will be described later.

Next, in step S4, the relationship between the estimated error matrix ΔP and the pseudo-error angle matrix _ΔΘp between the position estimated in step S3 and the true value for the current attitude information matrix q is calculated by, for example, a linear function ΔP=G( q) Approximate with ΔΘ _p +H (q) (equation (9)) to derive coefficient matrix G(q) and matrix H(q). These matrices are recorded in the storage unit 21 in step S5. Detailed processing of steps S4 and S5 will also be described later.

In step S6, it is determined whether or not m sets of data of the estimated error matrix ΔP dependent on the pseudo error angular matrix ΔΘ _p have been acquired. At this point, if m sets of data have not been acquired, the process returns to step S2, changes the pseudo error angle matrix ΔΘ _p in the internal arithmetic processing, and estimates the self-position again. At this time, in step S2 in the present embodiment, the value is repeatedly changed from the immediately preceding value by a step value equal to or less than the desired accuracy. By making changes in increments below the desired accuracy, it is possible to obtain the effect of increasing the resolution of changes in the output value (estimated error matrix ΔP in this configuration example) corresponding to the pseudo error angular matrix ΔΘ _p .

Generally, in the example of the position identification method, it is conceivable that the posture of the housing 300 changes. In the present embodiment, n sets of attitude-dependent data are acquired in order to derive the inter-sensor error angle matrix ΔΘ that also takes into account the data when the probe equipped with the self-localization system 100 is tilted. Therefore, in step S7, it is determined whether or not n sets of posture-dependent data have been acquired.

In this determination processing, if n sets of attitude-dependent data have not been acquired, the process returns to step S1 and the attitude of the housing 300 is changed by the gimbal 4 . Specifically, the drive mechanism control unit 40 outputs a command to change the posture by a predetermined angle, and drives the drive mechanism 41 to change the posture.

After changing the attitude, in the internal calculation processing in step S2, the pseudo error angle matrix ΔΘ _p is reset to 0 degrees for each axis, and the self-localization process is calculated. The number of these data is determined by whether or not a sufficient number of data can be obtained to represent the relationship between the estimated value and the true value of the estimated error matrix ΔP for the pseudo error angular matrix ΔΘ _p . Regarding the setting of the number of data, if the pseudo error angle matrix ΔΘ _p is in a linear relationship with the estimated error matrix ΔP, a large number of data is not required compared to the case of a strong nonlinear relationship. This number of data may be preset by the user using the self-localization system 100 .

In the example of FIG. 4, m sets of data depending on the pseudo-error angular matrix ΔΘ _p are obtained, and the linear function ΔP=GΔΘ _p +H is estimated based on these values.

On the other hand, since n sets of attitude-dependent data of the housing 300 are acquired, the final size of the coefficient matrix G and the matrix H indicating the relationship between the pseudo-error angle matrix _ΔΘp and the estimated error matrix ΔP is 2n×3. 2n×1. The finally obtained coefficient matrix G and matrix H are used in step S8 after the conditions are determined in steps S6 and S7.

In step S8, using the 2n × 3 coefficient matrix G and the 2n × 1 matrix H obtained so far, sensor Calculate the error angle matrix ΔΘ.

In the above description, an example of deriving the transformation matrix (coefficient matrix G(q) and matrix H(q)) and holding the data in steps S4 and S5 was described, but the timing is not limited to this. After it is determined that m sets of data have been acquired in S6, it may be configured to be executed. After acquiring the pairs, the transformation matrices for n pairs may be derived collectively.

FIG. 5 is a flow chart showing an example of the self-localization process (step S3) of FIG. In this example, time is acquired in step S11. This time is obtained from a clock 12 or the like provided in the system, or configured to be able to be obtained from the outside. In step S12 (state acquisition step), the gravitational vector acquired by the acceleration sensor 3, which is necessary for self-localization, and the attitude information matrix q for the inertial coordinate system I determined by the STT 5 are acquired. At this time, data from each sensor is acquired at a predetermined sampling time.

In step S13, filtering is performed to remove noise in the sensor data. If the sampling frequencies between the two sensors are different and downsampling is required, an aliasing filter is applied.

In step S14, using the obtained time, the attitude information matrix q of the STT coordinate system S with respect to the inertial coordinate system I, and the coordinate transformation matrix R ^c/A between the STT coordinate system S and the acceleration sensor coordinate system A, the acceleration sensor coordinate system A coordinate transformation matrix R ^c/A from A to the fixed coordinate system C of the star is derived based on equation (4).

Using this coordinate transformation matrix R ^{c /A} and the corrected rotation matrix R ^{A /A'} , the direction of gravity vector g ^A' obtained in the acceleration sensor coordinate system A' in step S15 is calculated as Convert to the fixed coordinate system C, and calculate the gravitational direction vector g ^c . At this time, when estimating the inter-sensor error angle matrix ΔΘ, the corrected rotation matrix R ^A/A′ is changed sequentially in step S2 as the result of No in step S4. In the case of pure position estimation using localization techniques (which is actually used to identify the position of the rover, which is its original purpose), the sensors estimated by a series of sensor-to-sensor error angle estimation techniques A matrix R ^A/A' obtained from the inter-angle error angle matrix ΔΘ is used in step S15. In step S16 (state estimation step), the position of the spacecraft on which the housing 300 is mounted is estimated using equations (1) and (2) using the gravitational vector for the fixed coordinate system of the star.

FIG. 6 is a flow chart showing an example of the transformation matrix derivation process in step S6 of FIG. In this example, in step S21 (true value acquisition step), a true value corresponding to the estimated value is acquired. In step S22 (estimated error derivation step), an estimated error matrix ΔP, which is an estimated error value between the estimated value and the true value, is calculated. In step S23, the coefficient matrix G and the matrix H are calculated using the estimated error matrix ΔP of the estimated values and the true values and the pseudo error angular matrix ΔΘ _p that was pseudo-varied in step S2 of FIG.

FIG. 7 is a flow chart showing an example of the procedure of the data holding process in step S5 of FIG. In this example, the coefficient matrix G(q) and the matrix H(q) calculated in step S23 of FIG. 6 are written in the storage unit 21 in step S31. In step S33, a combination of the coefficient matrix G(q) and the matrix H(q) recorded in the storage unit 21 in step S31 is output, and the decision processing in steps S6 and S7 of FIG. 4 is performed. At the time of passage, the 2n×3 coefficient matrix G and the 2n×1 matrix H are finally output. Note that when writing the coefficient matrix G(q) and the matrix H(q) to the storage unit 21 in step S31, it is assumed that the coefficient matrix G(q) and the matrix H(q) from the beginning to that point are combined. You can write.

2-2 System Configuration in Localization Method FIG. 1 is a diagram showing a configuration example of a self-localization system 100 of this embodiment. As shown in FIG. 2, in the present embodiment, the inter-sensor error angle matrix ΔΘ between the acceleration sensor 3 and the STT 5 is estimated by the inter-sensor error angle estimator 2, but as another aspect, the first Applicable for estimating the error angle between a sensor and a second sensor. In this configuration example, the sensor-to-sensor error angle matrix ΔΘ, the acceleration sensor 3, the STT 5, and the data acquired from the clock 12 are used to estimate the self-position by the method described above. The self-localization calculation device 1 is composed of a communication information processing section 10 , a self-position estimation section 11 , a clock 12 and a signal processing section 13 .

The self-localization arithmetic unit 1, as shown in FIG. A sensor-to-sensor error angle matrix ΔΘ, which is the deviation angle between the acceleration sensor and the ACC coordinate system A′ (the first sensor coordinate system after deviation), is estimated. At this time, the definition of the inter-sensor error angle ΔΘ is defined as the angle (deviation angle) at which the angle formed by the ACC coordinate system A with respect to the preset (calculated) STT coordinate system S changes to the ACC coordinate system A′. .

The acceleration sensor 3 outputs analog signals of acceleration in three axial directions that intersect each other and are ideally perpendicular to each other. It should be noted that the first sensor in other aspects is not limited to the acceleration sensor 3 as long as it can output the gravitational direction vector gA ^' .

The STT 5 includes an imaging sensor 50, an image processing unit 52 that processes acquired images, and an attitude determining unit 51 that calculates and determines the attitude of the STT coordinate system S with respect to the inertial coordinate system using data obtained from the images. . In another aspect, the second sensor may be a combination of a camera and an arithmetic processing unit capable of performing everything from image processing to attitude determination, instead of the STT. A processing system such as image processing may be included in the position identification calculation device 1 .

The signal processing unit 13 acquires data from the sensor at certain intervals and converts the data into digital values. In order to synchronize the sampling frequencies of the two sensors, if it is necessary to down-sample the data of one or both sensors, a low-pass filter is applied. Processing such as Butterworth filter and extended Kalman filter for reducing the influence of noise is also included.

The gimbal 4 includes a drive mechanism 41 for changing the attitude of the self-localization machine 200 and a drive mechanism control section 40 for controlling the drive mechanism 41 at predetermined timings. As long as the attitude of the self-localization machine 200 can be changed, the gimbal 4 including the drive mechanism 41 described above is not necessarily required, and any desired attitude changing means can be applied.

The transmitting/receiving section 6 includes a demodulator/modulator and filters for processing signals from the antenna 7 . Further, the communication information processing section 10 converts a signal to be received/transmitted into desired data. A method using relative VLBI, for example, can be employed to acquire the true position required for estimating the inter-sensor error angle. In this method, the self-localization device 200 transmits a specific radio wave from the antenna 7 to the earth station, and alternately acquires the signals transmitted from a plurality of antennas on the earth and the radio wave from a reference star, thereby achieving a high degree of accuracy. Accurate position determination (latitude and longitude on stars) can be performed. After the position is determined, the data is transmitted from the earth station to the spacecraft on the star. In order to acquire the true positional information transmitted from the earth station, the data is input to the processing section 20 included in the inter-sensor error angle estimating section 2 via the communication information processing section 10 . The processing unit 20 performs the arithmetic processing shown in FIG. 4 using this information. In FIG. 1, the communication information processing unit 10 is included in the self-localization calculation device 1, but a processing system that acquires position information and performs calculations can be combined as a separate system from the position identification calculation device 1. good too.

3 Modifications As described above, the target of the inter-sensor error angle estimation method of the present embodiment does not have to be the self-localization system, and the first sensor and the second sensor are limited to the acceleration sensor and the STT. can't In the above-described embodiment, the procedure for estimating the error angle between the acceleration sensor and the STT was shown using a self-localization system on a certain star as an example. Any system may be used as long as it estimates the state including the position of a moving object using a plurality of sensor information obtained.

1: self-localization arithmetic unit, 2: inter-sensor error angle estimation unit, 3: acceleration sensor, 4: gimbal, 5: star tracker (STT), 6: transmission/reception unit, 7: antenna, 10: communication information processing unit, 11: self-position estimation unit, 12: clock, 13: signal processing unit, 20: processing unit, 21: storage unit, 40: drive mechanism control unit, 41: drive mechanism, 50: imaging sensor, 51: attitude determination unit, 52: Image processing unit, 100: Self-localization system, 200: Self-localization machine, 300: Housing

Claims (7)

A method of estimating a sensor-to-sensor error angle between a first sensor and a second sensor, comprising:
A housing equipped with a first sensor that acquires a first state of a mounted housing and a second sensor that acquires a second state different from the first state of the housing. a posture change step for changing the posture;
a state acquiring step of acquiring the first state and the second state by the first sensor and the second sensor before and after the posture changing step;
a state estimation step of estimating a state of the housing based on the first state and the second state obtained in the state obtaining step;
a true value acquisition step of acquiring a true value corresponding to the estimated value estimated in the state estimation step;
an estimated error deriving step of setting a pseudo error angle between the first sensor and the second sensor and deriving an estimated error value between the estimated value and the true value;
and an inter-sensor error angle estimation step of deriving the inter-sensor error angle using the derived estimated error value. 2. The inter-sensor error angle estimation method according to claim 1, wherein said first sensor is an acceleration sensor and said first state is a gravitational vector. 3. The inter-sensor error angle estimation method according to claim 1, wherein said second sensor is a star tracker, and said second state is posture information of said housing. executing the posture changing step multiple times, and accordingly executing the state obtaining step multiple times;
setting the value of the pseudo error angle to a plurality of different values, repeatedly performing the estimated error derivation step a plurality of times;
4. The inter-sensor error angle estimation method according to claim 3, wherein the inter-sensor error angle is derived based on the plurality of estimated error values obtained thereby. a matrix calculation step of calculating a matrix representing a relationship between the estimated error value derived in the estimated error deriving step and the pseudo error angle used for deriving the estimated error value;
a storage step of storing the matrix in a storage unit;
In the inter-sensor error angle estimation step, based on the matrix stored in the storage unit, by calculating an error angle that makes an error between the true value and the estimated error value zero, the sensor 5. The inter-sensor error angle estimation method according to claim 1, wherein an inter-sensor error angle is derived. 6. The method for estimating an inter-sensor error angle according to claim 1, wherein said true value is acquired through communication means in said true value acquiring step. Position identification characterized by identifying the position of the housing based on the inter-sensor error angle estimated using the inter-sensor error angle estimation method according to any one of claims 1 to 6 Method.

Images