JP6091523B2 - Servo control device - Google Patents

Description

  The present invention relates to a servo control device.

  A machine tool is provided with a plurality of feed shafts, which are driven by a linear motor, a servo motor, or the like. At each feed axis, the position of the driven body is detected using a position detector so that the actual position of the driven body (table or the like on which the workpiece is fixed) matches the command position. Feedback control for correcting an error between the position of the driving body and the command position is performed. In feedback control, even if unknown disturbance is input, the driving force is controlled so as to cancel the disturbance. However, since the driving force corresponding to the error is input after detecting the error, the feed axis response is slow. There's a problem.

  The effect of frictional force, a kind of disturbance force, on the accuracy of contour motion is well known. For example, in the case of moving an arc locus using two axes orthogonal to each other in the XY plane, a sinusoidal motion command having a phase shifted by 90 degrees is given to each of the two axes. Then, at the point where the quadrant of the arc is switched, the movement direction of one of the feed axes is reversed. At this time, the direction of the friction torque and the frictional force generated at the contact portion such as a ball screw or a bearing, which is a component of the feed shaft, is also reversed, so that the control system of the reversing shaft responds with a certain delay. Therefore, a tracking error occurs in the response trajectory, and the actual trajectory passes slightly outside the command trajectory. This phenomenon is called quadrant protrusion and is a factor that reduces the accuracy of movement.

  Note that the frictional torque of the rotating system and the frictional force of the linear motion system can be equivalently converted by a constant determined by the configuration of the mechanical system. Therefore, in this specification, the friction torque and the frictional force are not distinguished from each other. In addition, the motor thrust of the linear motor and the motor torque of the rotary motor are not distinguished.

  As a conventional technique for solving the above problem, there is model-based feedforward control. In model-based feedforward control, a disturbance model is identified using disturbance force data measured in advance, the disturbance force at the time of reversing the motion direction is predicted, and a current command corresponding to the driving force that can overcome this disturbance force is issued. Add it to control the motor.

  For example, Patent Document 1 discloses a differential equation that divides a frictional force into three regions (a static friction region, a Coulomb friction region, and a viscous friction region) and expresses a relationship between speed and frictional force using predetermined parameters. A technique is disclosed in which a friction force is predicted by a model, a correction value that can overcome the friction force is calculated, and control is performed using the correction value. In Patent Document 1, the frictional force is modeled by a function only of speed.

  Further, for example, in Patent Document 2, in order to correct the hysteresis characteristic that the friction shows in the minute displacement region, the relationship between the displacement from the position where the movement direction of the driven body is reversed and the frictional force is modeled by a nonlinear function, A technique for correcting a frictional force by generating a correction value is disclosed. That is, in Patent Document 2, the frictional force is modeled by a function of only the position.

Japanese Patent Laid-Open No. 11-282506 Japanese Patent No. 4581096

  However, if the disturbance force is modeled using only one of the state quantities indicating the motion state (for example, only the velocity or the displacement from the motion direction reversal position), the model parameters are reset every time the motion conditions change. Cost. For example, when modeling with the function of only "speed", if the feed shaft is driven at a different driven body position, the total disturbance force fluctuates due to the influence of the disturbance force that depends on "position", so modeling error There is a problem that the correction does not function effectively.

  In addition, the position function in the above-mentioned prior art document 1 is strictly a function of a relative position from a position where the direction of motion is reversed, that is, a function of displacement. However, it is also known that there is a disturbance component that varies depending not only on the displacement but also on the position of the driven body relative to the origin of the machine. Therefore, in this specification, the position of the driven body with respect to the origin of the machine, that is, the global position is referred to as “position”, and is distinguished from the “displacement” amount from the movement direction reversal position.

  Also, when modeling with a function only of displacement from the reversal position (function of only “displacement”), if the feed shaft is driven at a different speed, it is completely affected by the disturbance force (friction force) depending on the speed. Since the disturbance force fluctuates, a modeling error occurs, and there is a problem that the correction does not function effectively.

  Thus, modeling using only one of the state quantities indicating the motion state is not sufficient for realizing a robust and highly accurate servo control device. Furthermore, the total disturbance force varies not only with the global position, the displacement from the moving direction reversal position and the speed of the driven body but also with the acceleration. Therefore, in a robust and more accurate servo control device, such a plurality of state quantities should be considered.

  The present invention has been made in view of the above, and has a disturbance correction function that eliminates the need to newly identify a parameter of a disturbance model even if each of the state quantities indicating the motion state of the driven body changes independently. An object is to obtain a robust and highly accurate servo control device.

In order to solve the above-described problems and achieve the object, the servo control device of the present invention performs control so that the position of the driven body driven by the motor including the position detector matches the position command value. A state quantity for estimating a plurality of state quantities of the driven body using a signal of the position detector provided in the motor and a signal of an acceleration sensor provided in the driven body, the servo control device Disturbance estimation in accordance with each of the plurality of state quantities by estimating each disturbance force using the estimation unit and each of the plurality of state quantities estimated and a plurality of disturbance models corresponding to each of the plurality of state quantities A disturbance estimation unit including an adder that outputs a value and generates a disturbance correction value from a linear sum of the disturbance estimation values of each of the plurality of state quantities, and adds the disturbance correction value to the signal from the position detector The motor A servo control unit that determines a drive current, and only a displacement-dependent disturbance force that is a displacement-dependent component of the disturbance force that varies due to a displacement of the driven body from a reversal position among the plurality of disturbance models is extracted. In the displacement dependent disturbance model, the displacement dependent disturbance force takes a steady value above the saturation value at which the displacement dependent disturbance force saturates with respect to the change in displacement from the reversal position, and the displacement dependent disturbance below the saturation value. This is a hysteresis-shaped model constructed by superimposing a plurality of models approximated by a plurality of functions whose force changes at a constant rate of change, and is a position that is a position-dependent component of disturbance force that varies depending on the position of the driven body. instead of depending disturbance force only extracted position-dependent disturbance model, when given repeatedly moving and stopping command to the servo controller in command width set, get Oite the stop position The data of the position-dependent disturbance force, using said table to save the relationship between the position and the position-dependent disturbance force of the driven body, wherein upon movement of the driven member, by referring to the table, the driven body It is characterized in that a correction command corresponding to the position is input.

  According to the present invention, even if each of the state quantities indicating the motion state of the driven body changes independently, a robust and highly accurate servo having a disturbance correction function that does not require the identification of a new disturbance model parameter. There is an effect that a control device can be obtained.

FIG. 1 is a block diagram illustrating configurations of the servo control device, the motor, and the driven body according to the first embodiment. FIG. 2 is a block diagram illustrating a configuration of the servo control unit according to the first to third embodiments. FIG. 3 is a block diagram of a configuration of the machine model unit according to the first embodiment. FIG. 4 is a flowchart illustrating the operation of the post-inversion displacement estimator according to the first and third embodiments. FIG. 5 is a block diagram illustrating a configuration of a disturbance estimation unit according to the first and second embodiments. FIG. 6 is a flowchart illustrating the operation of the disturbance estimation unit according to the first to third embodiments. FIG. 7 is a diagram illustrating a model in which a displacement-dependent disturbance force is approximated using a plurality of parallelogram models according to the first to third embodiments. FIG. 8 is a block diagram illustrating a configuration of the servo control device, the motor, and the driven body according to the second embodiment. FIG. 9 is a block diagram of a configuration of the state quantity estimation unit according to the second embodiment. FIG. 10 is a block diagram of the configuration of the servo control device, the motor, and the driven body according to the third embodiment. FIG. 11 is a block diagram of a configuration of a disturbance estimation unit according to the third embodiment.

  Embodiments of a servo control device according to the present invention will be described below in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a configuration of a first embodiment of a servo control device, a motor, and a driven body according to the present invention. FIG. 1 shows a servo control device 10, a motor 16 whose operation is controlled by the servo control device 10, a position detector 20 connected to the motor 16, and a driven body 18 driven by the motor 16. .

  The servo control device 10 includes a command value input unit 12, a servo control unit 14, a machine model unit 22, and a disturbance estimation unit 24.

  The command value input unit 12 outputs a position command to the servo control unit 14 and the machine model unit 22 according to the input target position of the driven body 18.

  The machine model unit 22 simulates a system including the servo control unit 14, the motor 16, and the driven body 18, and estimates the motion state amount of the driven body 18 based on the position command output from the command value input unit 12. Then, the estimated five state quantities are output to the disturbance estimation unit 24.

  The disturbance estimation unit 24 estimates a disturbance force from the five state quantities estimated by the machine model unit 22, and outputs the estimated disturbance force to the servo control unit 14 as a disturbance correction value.

  The servo control unit 14 performs feedback control using the position command from the command value input unit 12, the detector signal (detection position) from the position detector 20, and the disturbance correction value from the disturbance estimation unit 24. The movement of the driven body 18 is controlled by outputting a driving current (motor driving current) to the motor 16.

  FIG. 2 is a block diagram showing a configuration of the servo control unit 14. The servo control unit 14 includes a P (proportional) controller 30a, a PI (proportional integral) controller 32a, and a differentiator 34a. The position loop compensated by the P controller 30a and the PI controller 32a A velocity loop to be compensated.

  The P controller 30a outputs a speed command based on the position command from the command value input unit 12 and the detected position (detector signal) from the position detector 20.

  The PI controller 32a outputs a current command based on the speed command and the detected speed output from the P controller 30a. The detection speed is obtained by differentiating the detection position from the position detector 20 by the differentiator 34a.

  The current command output from the PI controller 32a is corrected by the disturbance correction value and output as a current command (motor drive current).

  FIG. 3 is a block diagram showing a configuration of the machine model unit 22. The machine model unit 22 includes a P controller 30b, a PI controller 32b, a torque constant multiplier 38, a feed axis inertia multiplier 40, an integrator 36a, an integrator 36b, a differentiator 34b, and an inversion A displacement estimator 42.

  The position command from the command value input unit 12 and the output of the integrator 36b are input to the P controller 30b.

  The output of the P controller 30b and the output of the integrator 36a are input to the PI controller 32b.

  The torque constant multiplier 38 receives the output of the PI controller 32b, and calculates and outputs the motor torque from the motor current command value.

  The feed shaft inertia multiplier 40 receives the output of the torque constant multiplier 38. The output of the feed axis inertia multiplier 40 is “acceleration” output from the machine model unit 22 as one of the state quantities. The inertia is calculated in advance.

  The output of the feed axis inertia multiplier 40 is input to the differentiator 34b. The output of the differentiator 34b is “jerk” output from the machine model unit 22 as one of the state quantities.

  The output of the feed axis inertia multiplier 40 is input to the integrator 36a. The output of the integrator 36a is a “speed” output from the machine model unit 22 as one of the state quantities.

  The output of the integrator 36a is input to the integrator 36b. The output of the integrator 36b is a “position” output from the machine model unit 22 as one of the state quantities.

  The output of the integrators 36a and 36b, that is, “speed” and “position” are input to the post-reversal displacement estimator 42. The output of the post-inversion displacement estimator 42 is “displacement” output from the machine model unit 22 as one of the state quantities.

  As described above, the machine model unit 22 calculates the “position”, “speed”, and “acceleration” of the feed shaft by simulating the feed shaft motion when the position command is input. Further, “jerk acceleration” is calculated from the rate of change of acceleration. Further, the post-reversal displacement estimator 42 calculates “displacement” from the motion direction reversal position using the velocity and the position.

  FIG. 4 is a flowchart showing the operation of the post-inversion displacement estimator 42. “Position” and “velocity” are input to the post-inversion displacement estimator 42. The post-reversal displacement estimator 42 determines whether or not the sign of the speed has been reversed (step S1). If it is determined that the sign of the speed has been reversed, it is determined that the direction of motion of the driven body 18 has been reversed. To do. Then, the position information at this time is stored in the memory as an inverted position, and the position information is updated. If it is determined that the sign of the speed is not reversed, the position information is not updated (step S2). The difference between the current position and the inverted position stored in the memory is output as a displacement from the inverted position (step S3).

  FIG. 5 is a block diagram showing a configuration of the disturbance estimation unit 24. The disturbance estimation unit 24 includes a jerk-dependent disturbance model 44, an acceleration-dependent disturbance model 46, a speed-dependent disturbance model 48, a displacement-dependent disturbance model 50, a position-dependent disturbance model 52, an adder 54, and a torque constant division. Instrument 56.

  FIG. 6 is a flowchart showing the operation of the disturbance estimation unit 24. The disturbance estimation unit 24 receives the five state quantities from the machine model unit 22 (step S11), and calculates a disturbance force value for each state quantity using a model for each state quantity (step S12). . For example, the jerk-dependent disturbance force is calculated using the jerk-dependent disturbance model 44. In this way, each disturbance force calculated according to each state quantity is input to the adder 54 and added to calculate the total disturbance force (step S13). Then, the driving force having the same magnitude as the total disturbance force calculated in this way is output to the servo control unit 14 as a disturbance correction value (step S14). The disturbance correction value is converted into a drive current value of the motor 16 and output.

  Here, each of the state quantity dependent disturbance models shown in FIG. 5 will be described.

In the jerk dependent disturbance model 44, the disturbance force is approximated by a constant with respect to the jerk for modeling. This is because the time during which the jerk-dependent disturbance force is applied is only a very short time when the acceleration of the driven body is changed, and there is no problem even if it is approximated by a constant while ignoring the fluctuation of the jerk. When the jerk-dependent disturbance force is F _j and the jerk is j, the jerk-dependent disturbance model 44 is expressed by Expression (1). sgn is a sign function.

In the acceleration-dependent disturbance model 46, the disturbance force is approximated by a linear expression with respect to the acceleration and is modeled. This is because the acceleration-dependent disturbance force is generated only when the driven body is accelerated, and the fluctuation of the acceleration-dependent disturbance force is proportional to the magnitude of the acceleration. When the acceleration-dependent disturbance force is F _a , the acceleration is a, the fluctuation rate of the disturbance force when the acceleration is changed is f _a1 , and the constant component not depending on the change in acceleration is f _a0 , the acceleration-dependent disturbance model 46 is expressed by an equation: It is expressed by (2).

In the velocity dependent disturbance model 48, the disturbance force is modeled by approximating the velocity with a polynomial. The velocity-dependent disturbance force is F _v , the velocity is v, the Coulomb friction force is F _c , the maximum static friction force is F _max , the maximum value of the displacement-dependent disturbance force is F _{x_max} , the constant that determines the Stribeck effect is α, the viscosity constant If D is D, the speed-dependent disturbance model 48 is expressed by Expression (3). By estimating the velocity-dependent disturbance force using the approximate equation obtained by subtracting the maximum value of the displacement-dependent disturbance force from the approximate equation of the Stribeck curve, the magnitude of the displacement-dependent disturbance force is prevented from being considered twice. ing. That is, the speed-dependent disturbance model 48 is a model configured by a function expressed by an equation obtained by subtracting the maximum value of the displacement-dependent disturbance force from a polynomial approximated by using a constant that determines the Stribeck effect.

In the displacement-dependent disturbance model 50, the relationship between the displacement and the disturbance force is approximated by a plurality of parallelogram functions and modeled. FIG. 7 is a diagram showing a model in which a displacement-dependent disturbance force is approximated by superimposing a plurality of parallelogram type functions. The parallelogram-type function shown in FIG. 7 shows that the displacement-dependent disturbance force takes a steady value above a saturation value at which the displacement-dependent disturbance force is saturated, and the displacement-dependent disturbance force changes at a constant value below the saturation value. It changes with rate. In the displacement dependent disturbance model 50, the hysteresis shape of the disturbance force is approximated by superimposing a plurality of parallelogram type functions, and the i-th parallelogram type disturbance model is called a loop i. When the disturbance force f _i, the displacement-dependent disturbance force and F _x, the displacement-dependent disturbance model 50 is expressed by the equation (4).

Here, when the displacement is x, the change rate of the disturbance force with respect to the displacement amount in the i-th parallelogram model is ΔF _i , and the steady value of the parallelogram model is F _{xi_max} , the disturbance force f _{xi of} each loop. Is approximated by equation (5).

Further, _assuming that the change rate of the displacement-dependent disturbance force immediately after the reversal of the motion direction is ΔF, ΔF _i and F _{xi_max} are determined so that the relational expressions of Expressions (6) and (7) are established. As the number of parallelogram models is increased, the hysteresis shape becomes a smooth non-linear shape. Therefore, i may be determined according to the modeling accuracy of the desired disturbance force.

  Note that the velocity-dependent disturbance force and the displacement-dependent disturbance force have a particularly large contribution to the total disturbance force. Therefore, the characteristics of the main disturbance force can be captured by the speed-dependent disturbance model 48 and the displacement-dependent disturbance model 50.

In the position-dependent disturbance model 52, modeling is performed with a function of a position approximated by least squares using a curve fitting method according to the relationship between the position of the feed axis measured in advance and the disturbance force. If the position-dependent disturbance force is F _g and the position-dependent disturbance force approximated by a polynomial is f _g , the position-dependent disturbance model 52 is expressed by Expression (8). The fluctuation frequency of the position-dependent disturbance force is generally a low frequency and often includes a periodic component such as a ball screw swing, so that it can be modeled by Fourier series or polynomial approximation. That is, the position-dependent disturbance model 52 may be a model of a function of the position of the driven body 18 approximated by a polynomial.

  The total disturbance force, which is the sum of the disturbance forces depending on the respective state quantities of the above formulas (1) to (8), is converted into the drive current value of the motor 16, and the disturbance correction value is output to the servo control unit 14. .

  That is, the servo control apparatus described in the present embodiment estimates a plurality of state quantities during the movement of the driven body from the position command value using a model simulating the control system, and outputs it to the disturbance estimation unit. A machine model unit is provided.

  As explained above, using a disturbance model that considers all changes in the state variables that determine the motion state, when the motion conditions change and the state amounts change independently, new optimal parameters are set. Even if it is not necessary, it is possible to appropriately correct the disturbance regardless of the motion condition, so that a robust and highly accurate servo control device can be realized.

  According to the present embodiment, since the machine model is used for estimating the state quantity, the state quantity can be estimated without being affected by the characteristics of the position detector.

Embodiment 2. FIG.
FIG. 8 is a block diagram showing the configuration of the second embodiment of the servo control device, the motor and the driven body according to the present invention. FIG. 8 shows a servo control device 10a, a motor 16a whose operation is controlled by the servo control device 10a, a position detector 20a connected to the motor 16a, and a driven body 18a driven by the motor 16a. An acceleration sensor 28 is provided on the driving body 18a.

  The servo control device 10a includes a command value input unit 12a, a servo control unit 14a, a state quantity estimation unit 26, and a disturbance estimation unit 24a.

  That is, the servo control device 10a shown in FIG. 8 includes a state quantity estimation unit 26 that simultaneously estimates a plurality of state quantities instead of the machine model unit 22, and includes an acceleration sensor 28 in the driven body 18a. This is greatly different from the servo control device 10 shown in FIG.

  The command value input unit 12a corresponds to the command value input unit 12 of the first embodiment, the servo control unit 14a corresponds to the servo control unit 14 of the first embodiment, and the motor 16a corresponds to the first embodiment. The driven body 18a corresponds to the driven body 18 of the first embodiment, the position detector 20a corresponds to the position detector 20 of the first embodiment, and the disturbance estimation unit 24a This corresponds to the disturbance estimation unit 24 of the first embodiment.

  Unlike the machine model unit 22 of the first embodiment, the state quantity estimation unit 26 estimates each state quantity using a detector signal from the position detector 20a and an acceleration sensor signal from the acceleration sensor 28. Therefore, even if the mass of the driven body 18a changes, the state quantity can be estimated without changing the inertia.

  FIG. 9 is a block diagram illustrating a configuration of the state quantity estimation unit 26. The state quantity estimation unit 26 includes differentiators 34c and 34d, a post-inversion displacement estimator 42a, and conversion coefficient multipliers 58a and 58b. The post-inversion displacement estimator 42a corresponds to the post-inversion displacement estimator 42 of the first embodiment.

  The detection position (detector signal) from the position detector 20a is input to the conversion coefficient multiplier 58b. The output of the transform coefficient multiplier 58b is a “position” output from the state quantity estimation unit 26 as one of the state quantities.

  The “position” that is the output of the transform coefficient multiplier 58b is input to the differentiator 34d. The output of the differentiator 34d is “speed” output from the state quantity estimation unit 26 as one of the state quantities.

  The inverted displacement estimator 42a receives the output of the conversion coefficient multiplier 58b and the output of the differentiator 34d. That is, “speed” and “position” are input. The output of the post-inversion displacement estimator 42a is “displacement” from the inversion position output from the state quantity estimation unit 26 as one of the state quantities.

  In this way, the post-inversion displacement estimator 42a can calculate the position, displacement, and speed. However, since the detector signal from the position detector 20a includes noise, if the second-order differentiation is performed using the detector signal from the position detector 20a to calculate acceleration, the reliability of the calculation result is low. There is a problem. Therefore, in the present embodiment, the acceleration sensor 28 provided on the driven body 18a is used.

  The conversion coefficient multiplier 58a receives the acceleration sensor signal from the acceleration sensor 28, and the output of the conversion coefficient multiplier 58a is “acceleration” output from the state quantity estimation unit 26 as one of the state quantities.

  The “acceleration” that is the output of the conversion coefficient multiplier 58a is input to the differentiator 34c. The output of the differentiator 34c is “jerk” output from the state quantity estimation unit 26 as one of the state quantities. Thus, the jerk is calculated from the rate of change of the acceleration sensor signal.

  As described above, according to the present embodiment, it is possible to estimate a plurality of state quantities using the detector signal of the position detector without using a machine model. In addition, if only the position detector is used, the noise included in the signal from the position detector affects the acceleration and jerk, and the reliability of the calculation result decreases. However, the acceleration sensor provided in the driven body Since the acceleration and jerk are calculated using the signals from, a plurality of state quantities input to the disturbance estimation unit can be estimated without reducing reliability.

  Further, since the state quantity is estimated without using the machine model, the state quantity can be estimated without changing the inertia even if the mass of the driven body changes.

  That is, the servo control device described in the present embodiment uses the position detector signal provided in the motor and the acceleration sensor signal provided in the driven body to move the driven body from the position command value. A state quantity estimating unit for estimating a plurality of state quantities at the time and outputting the estimated state quantity to a disturbance estimating unit;

  As described above, also in the present embodiment, a plurality of state quantities input to the disturbance estimation unit can be estimated as in the first embodiment. For this reason, when using a disturbance model that considers changes in all the state quantities that determine the movement state, the movement quantity changes independently as in a device that combines mechanical elements with different disturbance characteristics. In addition, since a disturbance can be appropriately corrected regardless of the motion conditions without setting a new optimum correction parameter, a robust and highly accurate servo control device can be realized.

Embodiment 3 FIG.
FIG. 10 is a block diagram showing the configuration of the third embodiment of the servo control device, motor and driven body according to the present invention. FIG. 10 shows a servo control device 10b, a motor 16b whose operation is controlled by the servo control device 10b, a position detector 20b connected to the motor 16b, and a driven body 18b driven by the motor 16b. .

  The servo control device 10b includes a command value input unit 12b, a servo control unit 14b, a machine model unit 22b, and a disturbance estimation unit 24b. The disturbance estimation unit 24b uses a simpler model than the disturbance estimation unit 24 of the first embodiment.

  That is, the servo control device 10b shown in FIG. 10 is significantly different from the servo control device 10 shown in FIG. 1 in that a simpler model than the disturbance estimation unit 24 of the first embodiment is used. In addition, the mechanical model unit 22 of the first embodiment outputs acceleration and jerk, but the mechanical model unit 22b of the present embodiment does not require a configuration for outputting acceleration and jerk.

  The command value input unit 12b corresponds to the command value input unit 12 of the first embodiment, the servo control unit 14b corresponds to the servo control unit 14 of the first embodiment, and the motor 16b corresponds to the first embodiment. The driven body 18b corresponds to the driven body 18 of the first embodiment, the position detector 20b corresponds to the position detector 20 of the first embodiment, and the disturbance estimation unit 24b This corresponds to the disturbance estimation unit 24 of the first embodiment.

  In the first and second embodiments, the disturbance estimation units 24 and 24a estimate the disturbance force with high accuracy using five state quantities. However, a feed shaft with low reproducibility of disturbance force due to low assembly accuracy and machine element accuracy is not effective even when a high-precision disturbance model is used. Furthermore, time is wasted in building unnecessary models. Therefore, in the present embodiment, a state quantity that has a particularly large influence on the fluctuation of the total disturbance force is selected.

  FIG. 11 is a block diagram showing a configuration of the disturbance estimation unit 24b in the present embodiment. The disturbance estimation unit 24b includes a speed-dependent disturbance model 48b, a displacement-dependent disturbance model 50b, a position-dependent disturbance model 52b, an adder 54b, and a torque constant divider 56b.

First, the speed-dependent disturbance model 48b is simplified as compared with the first embodiment, and is expressed by Expression (9). In other words, the speed-dependent disturbance model 48b is configured by a function in which the speed-dependent disturbance force is expressed by a first-order term of velocity having a viscosity as a proportional constant and a constant term of a difference between the maximum value of the Coulomb friction force and the displacement-dependent disturbance force. Model. It is _assumed that the speed-dependent disturbance force is F _v , the speed is v, the Coulomb friction force is F _c , the maximum value of the displacement-dependent disturbance force is F _{x_max} , and the viscosity constant is D.

  This is because the speed-dependent disturbance force exhibits the most specific behavior in the low-speed region from the stop, and therefore it is sufficient to model only the low-speed region from the stop in an apparatus with low reproducibility of the disturbance force. In this way, by modeling only the low speed region from the stop, it is possible to omit the trouble of identifying unnecessary parameters.

  The displacement-dependent disturbance model 50b is expressed by Expression (10) using the approximate expression proposed by Dahl. That is, the displacement-dependent disturbance model 50b is a model configured by a function obtained by approximating the relationship between the displacement from the inversion position and the displacement-dependent disturbance force by a polynomial.

  According to equation (10), the relationship between displacement and displacement-dependent disturbance force can be approximated by a polynomial, which is effective when the relationship between displacement and displacement-dependent disturbance force cannot be handled by the parallelogram model.

  The position-dependent disturbance model 52b is not modeled by a function, but the position-dependent disturbance force is obtained by referring to the relationship between the position measured in advance and the disturbance force. For example, the disturbance force is measured in the entire region of the feed axis while stopping the driven body 18b every 1 mm, and the position-dependent disturbance force data at a predetermined position is acquired, and the relationship between the position and the position-dependent disturbance force is used as a table. What is necessary is just to set it as the structure which preserve | saves at memory and refers this data table at the time of the exercise | movement of the to-be-driven body 18b. That is, in this embodiment, the position-dependent disturbance force is estimated by referring to a data table of the position and position-dependent disturbance force measured in advance instead of the position-dependent disturbance model.

  In this embodiment, the disturbance force is estimated using the speed-dependent disturbance model, the displacement-dependent disturbance model, and the position-dependent disturbance model, but the present invention is not limited to this. For estimation of the disturbance force, at least a speed-dependent disturbance model and a displacement-dependent disturbance model may be used. This is because the disturbance force represented by these two models has a particularly large influence on the total disturbance force.

  That is, in the servo control device described in the present embodiment, at least the plurality of state quantities used by the disturbance estimation unit are displacement and speed from the inversion position, and the disturbance model used by the disturbance estimation unit is driven A displacement-dependent disturbance model in which only the displacement-dependent component of the disturbance force is extracted, which varies depending on the displacement from the inversion position of the body, and a velocity dependency in which only the velocity-dependent component of the disturbance force, which varies with the speed of the driven body What is necessary is just to be comprised by the disturbance model.

  In the present embodiment, the state quantity used by the disturbance estimation unit is estimated by the machine model unit, but the present invention is not limited to this. A state quantity estimation unit may be used as in the second embodiment.

  In this embodiment, the disturbance estimation unit estimates the disturbance force using a displacement-dependent disturbance model, a speed-dependent disturbance model, and a position-dependent disturbance model. However, the present invention is not limited to this, and the acceleration-dependent disturbance is further limited. A model may be used.

  As described above, according to the present embodiment, since the configuration of the disturbance estimation unit can be simplified as compared with the first embodiment, the disturbance force is reduced in the feed shaft with low disturbance force reproducibility. Only the characteristics that have a large influence on the fluctuations of the model are modeled, and it is possible to efficiently calculate the disturbance correction value without wasting time for constructing an unnecessary model.

  As described above, the servo control device according to the present invention is useful for machine tools and the like that require high-precision machining.

10, 10a, 10b Servo control device, 12, 12a, 12b Command value input unit, 14, 14a, 14b Servo control unit, 16, 16a, 16b Motor, 18, 18a, 18b Driven body, 20, 20a, 20b Position Detector, 22, 22b Mechanical model section, 24, 24a, 24b Disturbance estimation section, 26 State quantity estimation section, 28 Acceleration sensor, 30a, 30b P controller, 32a, 32b PI controller, 34a, 34b, 34c, 34d Differentiator, 36a, 36b Integrator, 38 Torque constant multiplier, 40 Feed axis inertia multiplier, 42, 42a Inverted displacement estimator, 44 jerk dependent disturbance model, 46 acceleration dependent disturbance model, 48, 48b speed dependent disturbance Model, 50, 50b Displacement dependent disturbance model, 52, 52b Position dependent disturbance model, 54, 5 b adder, 56 and 56B torque constant divider, 58a, 58b transform coefficient multipliers, S1 to S3, S11 to S14 steps.

Claims (7)

A servo control device that performs control so that a position of a driven body driven by a motor including a position detector matches a position command value,
A state quantity estimating unit that estimates a plurality of state quantities of the driven body using a signal of the position detector provided in the motor and a signal of an acceleration sensor provided in the driven body;
Estimating each disturbance force using each of the estimated plurality of state quantities and a plurality of disturbance models corresponding to each of the plurality of state quantities, and outputting a disturbance estimated value corresponding to each of the plurality of state quantities. A disturbance estimation unit including an adder that generates a disturbance correction value from a linear sum of the disturbance estimation values of each of the plurality of state quantities;
A servo control unit that determines the drive current of the motor by adding the disturbance correction value to the signal from the position detector;
Of the plurality of disturbance models, a displacement-dependent disturbance model in which only a displacement-dependent disturbance force that is a displacement-dependent component of the disturbance force that varies due to the displacement of the driven body from the inversion position is extracted is a displacement from the inversion position. The displacement-dependent disturbance force takes a steady value above the saturation value at which the displacement-dependent disturbance force is saturated with respect to the change in the value, and the displacement-dependent disturbance force changes at a constant rate of change below the saturation value. It is a model of hysteresis shape configured by overlapping multiple approximate models,
Instead of a position-dependent disturbance model in which only the position-dependent disturbance force that is a position-dependent component of the disturbance force, which fluctuates depending on the position of the driven body, a move and stop command is sent to the servo controller with a set command width. Repeat gave upon, the data of the position-dependent disturbance force obtained Oite the stop position, using the stored the relationship between the position and the position-dependent disturbance force of the driven body table, the driven body A servo control device, wherein a correction command corresponding to a position of the driven body is input with reference to the table during exercise. Oite to claim 1,
The plurality of state quantities used by the disturbance estimation unit further includes acceleration of the driven body,
The plurality of disturbance models used by the disturbance estimation unit are:
A servo control device further comprising an acceleration-dependent disturbance model in which only an acceleration-dependent disturbance force that is an acceleration-dependent component of a disturbance force that varies depending on the acceleration of the driven body is extracted. In claim 2 ,
The plurality of state quantities used by the disturbance estimation unit further includes jerk of the driven body,
The plurality of disturbance models used by the disturbance estimation unit are:
A servo control device further comprising a jerk-dependent disturbance model in which only a jerk-dependent disturbance force that is a jerk-dependent component of the disturbance force, which fluctuates due to the jerk of the driven body, is extracted. In claim 2 or claim 3 ,
The speed-dependent disturbance model, in which only the speed-dependent disturbance force, which is the speed-dependent component of the disturbance force that fluctuates depending on the speed of the driven body, is extracted from the polynomial approximated using a constant that determines the Stribeck effect, A servo control device characterized by being a model constituted by a function expressed by an expression obtained by subtracting the maximum value of disturbance force. In claim 2 or claim 3 ,
The speed-dependent disturbance model in which only the speed-dependent disturbance force, which is the speed-dependent component of the disturbance force that fluctuates according to the speed of the driven body, is extracted. And a servo control device characterized in that the model is constituted by a function expressed by a constant term of a difference between the maximum value of the Coulomb friction force and the displacement-dependent disturbance force. In claim 2 or claim 3 ,
The servo control device according to claim 1, wherein the acceleration-dependent disturbance model is a model obtained by approximating the acceleration-dependent disturbance force with respect to the acceleration by a linear expression. In claim 3 ,
2. The servo control device according to claim 1, wherein the jerk dependent disturbance model is a model obtained by approximating the jerk dependent disturbance force with respect to the jerk by a constant.

Images