TW200928629A - Motor controller - Google Patents

Abstract

To allow a motor with a connected load to be subjected to control gain adjustment safely and quietly without becoming vibratory. A motor controller includes a mechanical constant identifier (121) that receives a response V for frequency response measurement and outputs a mechanical constant, a stability boundary calculator (122) that receives the mechanical constant for gain adjustment and outputs a stability boundary, a calibration curve calculator (123) that receives the stability boundary and outputs a calibration curve, an initial value setter (124) that receives the calibration curve and outputs a control gain initial value Kin, and a control gain adjuster (125) that receives the calibration curve and the control gain initial value Kin and outputs a compensation amount Kf and a feedback control gain Kvj.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motor control device that adjusts a control gain to control an operation of a motor after a load is connected. [Prior Art] A conventional motor control device minimizes the positional deviation evaluation function of the positional deviation of the positional deviation at the time of transition, and sets the position proportional control gain and the speed proportional control gain to a steady state. The error self-sufficient area of the torque command does not exceed the threshold 値 (see, for example, Patent Document 1). FIG. 5 is a block diagram showing an example of a conventional motor control device. In Fig. 5, the figure number 501 is the target command unit during gain automatic adjustment, the figure number 52 is the position proportional control gain, the figure number 503 is the speed ratio control gain, the figure number 504 is the torque control part, and the figure number 505 is the encoder. Figure 506 φ is the motor, Figure 507 is the ball screw mechanism, Figure 508 is the position deviation evaluation function calculation unit, Figure 5 09 is the position deviation evaluation function memory, and Figure 510 is the evaluation function minimum gain determination unit. Reference numeral 511 is a torque command evaluation function calculation unit, reference numeral 512 is a torque command evaluation function storage unit, and reference numeral 5 13 is a torque command determination unit. The target 値 command unit 50 1 outputs the position command Xr at the time of automatic gain adjustment. The position proportional control gain 502 is a position following deviation Ex and a position proportional control gain setting 値Kp after inputting the position data Xf from the position command Xr, and then outputting the position following deviation Ex amplified position proportional control gain -4-200928629 setting The signal after 値Κρ times is the speed command νΓ. The speed proportional control gain 503 inputs the speed following deviation Ev and the speed proportional control gain setting 値Κν after subtracting the speed data Vf from the speed command Vr, calculates the torque command Tref, and then calculates the torque command unit 504 and the torque command evaluation function. The part 511 outputs. The torque control unit 504 inputs the torque command Tref and the encoder pulse, calculates the position data Xf and the speed data Vf, and returns the motor current to the motor 506. The encoder 505 detects the position of the motor 506 and outputs it to the torque control unit 504 as the above encoder pulse. The motor 506 inputs the motor current Im to drive the ball screw mechanism 5 07 to which it is coupled. The positional deviation evaluation function calculation unit 508 outputs the positional deviation deviation Ex, and then outputs the positional deviation evaluation function, which is the error multiplication area of the positional follow-up deviation Ex at the time of transition, to the positional deviation evaluation function storage unit 509.位置 The positional deviation evaluation function storage unit 509 inputs the positional deviation evaluation function and then stores the input signal as the positional deviation evaluation function, and outputs it to the evaluation function minimum gain determination unit 510. The evaluation function minimum gain determining unit 510 inputs the above-described positional deviation evaluation function memory 値 and then the combination of the control gains that minimize the input signal, that is, the evaluation function minimum gain is output to the torque command determining unit 513. The torque command evaluation function calculation unit 511 inputs the torque command Tref and outputs the torque command evaluation function, that is, the torque command evaluation function of the torque command Tref at the steady state, to the torque command evaluation function storage unit 512. -5- 200928629 The torque command evaluation function storage unit 512 inputs the torque command evaluation function and then stores the input signal as the torque command evaluation function memory to the torque command determination unit 5 1 3 . The torque command determining unit 5 1 3 inputs the evaluation function minimum gain and the torque command evaluation function memory 値, and when the torque command evaluation function does not exceed 闽値, the evaluation function minimum gain is set as the position proportional control gain 値Kp The position proportional control gain 502 is outputted, and φ is outputted as the speed proportional control gain setting 値Κν to the speed proportional control gain 503 as the speed of the evaluation function minimum gain. As described above, the conventional motor control device minimizes the error deviation self-occupying area, that is, the positional deviation evaluation function at the time of transition, and sets the position proportional control gain and the speed proportional control gain to the steady state torque. The error self-sufficiency area of the command does not exceed 闽値. [Patent Document 1] Japanese Laid-Open Patent Publication No. Hei 6-208404 (page 7 and Fig. 1) ❹ [Summary of the Invention] The conventional motor control device is configured to shift the position at the time of transition. The error self-multiplied area, that is, the position deviation evaluation function becomes the minimum, and the error proportional multiplication area of the torque command when the position proportional control gain and the speed proportional control gain are set to the steady state does not exceed the threshold 値, so the control system includes the observer and the filter When the control unit including the control object has a stable control gain and the stable region becomes a complicated shape, the load connected to the motor -6-200928629 may cause vibration or instability, so that there is a problem that control cannot be performed. . The present invention has been made in view of the above problems, and it is possible to calculate a control gain stabilization region while maintaining the control gain in the above-described manner, so that the operation of the load connected to the motor does not become unstable, and the adjustment can be quietly and in a short time. Motor control q for controlling gain [Means for Solving the Problem] In order to solve the above problems, the present invention is constituted as follows. According to the invention of the first aspect of the invention, the feedback controller (Vr) and the control target response (V) are matched to each other to control the pre-interference correction torque command (Τα); The command (Τα) subtracts the disturbance presumption 値 (the torque command (Tref) and the above response (V), calculates the disturbance observer T (Td) interference observer; the input frequency response measurement 〇 (Fr) and the torque command (Tref In the frequency response measurement, when the rate response measurement signal (Fr) is used as the current command (lr), the torque command (Tref) is used as the current command (output signal converter; according to the above current command (Ir) a torque control unit that drives the motor current motor current (Im) and a vibration control unit that calculates a vibration signal (Vs) based on the response (V) and the torque command (Tref), and constitutes vibration control for the command (Vr) The motor control device of the signal (Vs) is characterized in that the feedback of the feedback controller is calculated based on the response (V) to provide an intra-area vibration or device. Calculate, the data from the above Td) after the above-mentioned dry signal is used to transmit the above controller; the dynamic control is added, with gain (-7-200928629 KVj) and the above control amount (Kf) control gain adjustment In the invention according to the first aspect of the invention, the frequency response measurement signal (Fr) is a frequency sine wave signal. The invention according to the invention of claim 1 is characterized in that the frequency response measurement signal (Fr) is formed by a complex frequency sine wave. According to the invention of the first aspect of the invention, the control gain adjustment unit is configured to: according to the response (V) in the frequency response measurement a mechanical constant identifier for identifying a mechanical constant of the control object; and calculating a closed loop system for including the feedback controller and the control target based on the mechanical constant The feedback control gain and the contour line of the correction amount (Kf) stable region, that is, the stability boundary operator of the stability boundary; and the adjustment curve capable of adjusting the feedback control gain and the upper reference correction amount (Kf) according to the stability boundary An adjustment curve calculator for improving the closed-loop responsiveness and the interference suppression characteristic while maintaining the stability of the control target; and calculating the feedback gain (KVj) and the correction amount (Kf) of the control gain adjustment based on the adjustment curve That is, an initial clamper that controls the gain initial 値 (Kin); and a control gain adjuster that adjusts the feedback control gain (KVj) and the correction amount (Kf) based on the adjustment curve and the control gain initial 値 (Kin). The invention according to claim 4, wherein the mechanical constant -8-200928629 identifier calculates a frequency response of the control target according to the frequency. The above mechanical constant is calculated in response. The invention according to claim 4, wherein the stable boundary line operator calculates the equivalent unit feedback system of the closed loop system based on the mechanical constant. The stability margin is calculated by setting the phase margin of the equivalent unit feedback system to a positive phase margin. In the invention described in claim 6, the phase margin setting 设定 is set such that the mechanical constant is uncertain. The above-mentioned uncertainty is allowed in the situation to make the closed loop system stable. In the invention described in the fourth aspect of the invention, the adjustment curve computing unit uses the above correction amount when the stability curve is one. Kf) is zero, and the feedback control gain (Kvj) is a straight line which can increase the control gain (値, which is a point at which the correction amount (Kf) is zero to the stability curve, and is increased as the above-mentioned adjustment curve output. When the number of the two curves is "the curve which becomes the intersection of the two stable curves, which is the intersection of the two stable curves in the middle of the two stable curves", the above-mentioned adjustment curve is output. The invention of the invention of claim 8 is characterized in that the initial setting device is "the point on the adjustment curve and from the control gain" The feedback control gain (Kvj) and the above-mentioned 200928629 correction amount (Kf) of the point at which the optimum 値 is set to the set amount are output as the control gain initial 値 (Kin). The invention described in claim 9 is the invention according to claim 9, wherein the control gain adjuster is configured to follow an adjustment curve from an initial stage of the control gain ( Ku) increases the above-described feedback control gain (Kvj) and the above-described correction amount (Kf) toward the above-described control gain optimum and then outputs it. 〇 [Effect of the Invention] According to the invention described in the first to seventh aspects of the patent application, the stable region of the control gain can be calculated, and the control gain can be adjusted in the stable region, so that the control object does not vibrate. The control gain can be adjusted quietly and safely, and the automatic and high control performance can be realized. According to the invention described in the eighth to tenth claims of the patent application, the control gain is the least likely to be caused in the stable region. The vibration adjustment curve is adjusted from the initial control gain close to the control gain, so that the control object does not vibrate, the control gain can be adjusted quietly, safely, and in a short time, and automatic and high control can be realized. [Embodiment] [Best Embodiment of the Invention] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The actual motor control device contains various functions or means, but only the functions or means for explaining the present invention are described in the figure - -10-200928629. In addition, the same name is used to omit the same figure number as much as possible. [Embodiment 1] Fig. 1 is a block diagram showing a schematic configuration of a motor control device according to a first embodiment of the present invention. In Fig. 1, the figure 101 is the command generator figure number, the figure 102 is the feedback controller, the figure 103 is the input signal converter, the figure 0 104 is the torque controller, and the figure 105 is the control object, the figure number 106 is a response detector, reference numeral 107 is an interference observer, reference numeral 108 is a frequency response measurement signal generator, picture number 110 is a vibration suppression unit, picture number 120 is a control gain adjustment unit, and picture number 111 is a vibration component operation The figure number 112 is the phase adjuster, the figure 113 is the correction amount adjuster, the figure 121 is the mechanical constant identifier, the figure 122 is the stability boundary operator, the figure 123 is the adjustment curve operator, and the figure 124 is The initial setter, figure 125, is the control gain adjuster. 〇 The control object 105 is a motor to which a load is connected. The command generator 1 〇 1 is the output command Vr. The response detector 106 detects the response of the control object 105 and then outputs a response V. The frequency response measurement signal generator 108 is an output frequency response measurement signal Fr. The feedback controller 102 inputs the vibration suppression follow-up deviation after subtracting the response v from the command Vr and adding the vibration suppression signal Vs, and the feedback control gain KVj, and then calculates the disturbance correction pre-torque command T a . -11 - 200928629 The input signal converter 103 inputs the torque command Tref after subtracting the interference estimation 値Td from the disturbance correction torque command Τβ, and the frequency response measurement signal Fr, and then uses the torque command Tref as the torque adjustment. The current command Ir is input to the torque controller 104, and is input to the torque controller 104 by the frequency response measurement signal Fr as the current command Ir at the time of frequency response measurement. The torque controller 104 inputs a current command Ir to move the motor current Im to the motor of the control object 105 to thereby drive the motor. The disturbance observer 107 inputs the torque command Tref and the response V at the time of gain adjustment and then calculates the interference estimation 値Td. The vibration suppressor 110 is composed of a vibration component arithmetic unit 11, a phase adjuster 112, and a correction amount adjuster 113. At the time of gain adjustment, the torque command Tref, the response V, and the correction amount Kf are input, and the vibration suppression signal Vs is calculated. The vibration component computing unit 111 inputs the torque command Tref and the response V, and then calculates the vibration component of the response V to output to the phase adjuster 112. The phase adjuster 112 inputs the above-described vibration component, and then calculates a phase adjustment vibration component obtained by adjusting the phase of the input signal to the correction amount adjuster 113. The correction amount adjuster 113 inputs the phase adjustment vibration component and the correction amount Kf, and then calculates the vibration suppression signal Vs obtained by multiplying the phase adjustment vibration component by the correction amount Kf. The control gain adjustment unit 120 is composed of a mechanical constant identifier 121, a stability boundary operator 122, an adjustment curve operator 123, an initial parameter setter-12-200928629124, and a control gain adjuster 125, and is input when the frequency is measured. In response to V, at the time of gain adjustment, the correction amount Kf and the feedback control gain described above are calculated. The mechanical constant identifier 1 2 1 inputs the response V to the frequency response measurement, and then calculates the mechanical constant of the control object 105 to be output to the stabilization boundary operator 122. Specifically, the mechanical constant identifier 121 calculates a frequency response (Bode diagram) based on the response V, and calculates an anti-resonance frequency and a resonance frequency from the frequency of the groove of the amplitude map and the peak of the induced peak, from the groove and the above Q値 of the peak calculates the attenuation coefficient of the anti-resonance and the attenuation coefficient of the resonance, and calculates the moment of inertia of the control object 105 based on the amplitude of the amplitude map which is relatively lower than the lowest frequency among the anti-resonance frequencies, and the anti-resonance is performed. The frequency, the resonance frequency, the attenuation coefficient, and the moment of inertia are used as the mechanical constant. The stability limit line operator 122 inputs a mechanical constant at the time of gain adjustment, and calculates a closed-loop system including the feedback controller 1 〇2, the control object 1 〇5, the response detector 106, and the disturbance observer 107 as a critically stable feedback control. The curve of the combination of the gain Kvj and the correction amount Kf is a stability boundary, and is then output to the adjustment curve operator 123. In the gain adjustment, the adjustment curve operator 123 inputs the above-described stable boundary line 'the adjustment curve formed by the combination of the feedback control gain KVj and the correction amount Kf in which the response V does not become the vibration in the stable region surrounded by the stable boundary line, and then outputs to the adjustment curve The initial setter 124 and the control gain adjuster 125 ° -13 - 200928629 The initial setter 124 inputs the above-mentioned adjustment curve during the gain adjustment, and calculates the point at which the response V is optimal when the adjustment curve is relatively close to the normal operation. The point at which the gain point is controlled is outputted to the control gain adjuster 125 at this point as the control gain initial 値Kin. The control gain adjuster 125 inputs the adjustment curve and the control gain initial 値Kin at the gain adjustment, and changes the correction amount Kf and the feedback 〇 control gain KVj from the control gain initial 値Kin along the adjustment curve toward the optimal control gain point, The correction amount Kf is output to the correction amount adjuster 113, and the feedback control gain KVj is output to the feedback controller 102. The frequency response measurement signal Fr is not limited to any signal as long as it can measure the frequency response of the motor control device, but the sweep sine wave signal is preferred. In addition, a signal composed of a plurality of frequency sine waves can also be used. When the frequency response measurement signal Fr is used as the frequency sine wave signal, the frequency response of the control object 105 can be accurately detected in a short time. Further, when the frequency response measurement signal Fr is a signal composed of a complex frequency sine wave, the frequency response of the control object 105 can be accurately detected in a short time. The present invention differs from the prior art in that it has a portion having a configuration in which a mechanical constant identifier 121 that outputs a mechanical constant after inputting a response during frequency response measurement is provided, and the mechanical input is input during gain adjustment. a constant boundary operator 122 for outputting a stable boundary; an adjustment curve operator 123 for outputting the adjustment curve after inputting the stable boundary; and an initial setter 1 24 for outputting the control gain initial 値Kin after inputting the curve; and inputting the above adjustment curve And a control gain adjuster 125 that outputs a correction amount Kf and a feedback control gain KVj after controlling the gain initial 値Kin. -14-200928629 Hereinafter, the detailed configuration in which the control gain adjustment unit 120 calculates the correction amount Kf and the feedback control gain KVj is disclosed. In this embodiment, the feedback controller 102 is combined with the speed PI controller in a series of low-pass filters, that is, torque command filters, and the control object 105 includes a notch filter, dead time and three inertia. mechanism. Further, the command Vr is a speed command, and the response V is a motor speed (hereinafter referred to as "motor speed V"). Q speed PI controller transfer function Ge(s), torque command filter transfer function Gt(s), notch filter transfer function Gn(s), three inertia mechanism transfer function Gp(s), and band The pass filter, that is, the transfer function Gb(s) of the phase adjuster 112, is expressed by equations (1) to (5), respectively. !:+y (i) \ 1iS) Only KVj is the speed proportional control gain, and Ti is the speed control integral time constant. The 〇 speed proportional control gain KVj and the speed control integral time constant Ti belong to the feedback control gain, but it is known that the speed control integral time constant Ti can be determined subordinately by determining the speed proportional control gain Kvj. Therefore, the feedback control gain of the present embodiment is only for A method of calculating the speed proportional control gain KVj will be described.

Only Tf is the torque command filter time constant. -15- 200928629 Only, ωη is the notch filter frequency, tn is the notch filter attenuation coefficient

Only J is the total moment of inertia of the three-inertia mechanism, ω rl is the first resonance frequency, Γη is the first resonance attenuation coefficient 'ω3ΐ is the first anti-resonance frequency, ral is the first anti-resonance attenuation coefficient' 〇r2 is the second The resonance frequency, 0 is the second resonance attenuation coefficient, ω32 is the second anti-resonance frequency, and (a2 is the second anti-resonance attenuation coefficient.

(r-5 + l)? (5) Only r is the pass filter time constant. The disturbance observer 1〇7 estimates the interference caused by the gravity of the control object 105 and the interference caused by gravity interference or clutter. The transfer function of the disturbance observer 107 is expressed by the equation (6) ^

Only T/ is the Laplace transform of the interference estimate 値Td, J〇 is the nominal moment of inertia, V is the Laplace transform of the normalized estimated interference D, and ^ is the Laplace transform of the motor speed V, Tref4 It is a Laplace transform that allows the control target 105 to constitute a torque command Tref generated by the torque driven by the motor, and Gvd(s) is a transfer function from the motor speed < to the disturbance estimation 値T/, 1!, 12 is The disturbance observer gain, Gtd(s), is the transfer function from the torque command Tre to the disturbance estimate 値T/. -16- 200928629 The vibration component calculator 111 calculates the motor speed after the vibration component is removed by the speed observer disclosed in the equation (7), and calculates the motor speed v from the motor speed v M to remove the motor component after the vibration component is removed. Vibration component.

Only, Ve* is the Laplace transform for estimating the motor speed Ve. 'Gtv(S) is the transfer function from the torque command Tref* to the estimated motor speed Ve·' is the speed observer gain, and GVV(S) is from The transfer function from the motor speed V + to the estimated motor speed V / . The phase adjuster 112 is constructed by a band pass filter disclosed in the formula (5), and calculates the phase adjustment vibration component that has been changed so that the phase of the input vibration component coincides with the phase of the command V. Fig. 2 is a view showing an equivalent unit feedback system of the motor control device according to the first embodiment of the present invention shown in Fig. 1. In Fig. 2, reference numeral 201 is an open loop transfer function. Fig. 2 is a view obtained by converting the portion excluding the control gain adjustment unit 120 of Fig. 1 into a unit feedback system. The open-loop transfer function 201 is a first rational transfer function Gi(s) shown by the dead time VIII and (8), and a second rational transfer function G2(s) represented by the formula (9), and (10) Said.

(8) Only Kf is the correction amount of Fig. 1. -17- (91 200928629

Since the stable boundary line is calculated by the stability boundary operator 122, the open-loop transfer function 201 of Fig. 2 is approximately as shown in the equation (1).

[,琢oik A>/.十·A* (TO)

Only Νσ1ν, Dq1Q and D. ^ is a polynomial of the Laplace variable s independent of the velocity proportional control gain Kvj and the correction amount Kf. The approximation from the irrational transfer function to the rational transfer function is using the Pade approximation ° (Kvj, Kf) plane The closed loop system is a stable stable region which is calculated from the condition that the phase margin of the open loop transfer function 201 is positive in the feedback system. The condition for calculating the phase margin is to be substituted into (10), and thus the equation (11) can be obtained. Only j is an imaginary unit. Only Nr, Ni, D〇r, Dvr, Dgi, and DVi are polynomials associated with the Laplacian variable s independent of the speed proportional control gain KVj and the correction amount Kf. In the gain crossover frequency ω·: The amplitude of the function 2〇1 is 1 and the equation (12) can be established. -18- 200928629

The formula (13) can be obtained by deforming the formula (12). Υ _ ·^ · + ^rpGf' +3⁄4 · (3⁄4 + 『f ii+jf ~~^ (13) Ο

When Kvj · Kf's decision is made, KVj can be calculated using (1 3 ). Next, the condition that the phase margin of the open-loop transfer function 201 becomes 5 is expressed by the formula (14). ^ |14) Gain crossover frequency ω. The phase of the open loop transfer function 201 is expressed by the equation (15). M々尾.(3⁄4 +3⁄4.~为私单柄+%今Aj ci S3 _ hnra ditch+马·9 iron D0 +Kf · Κν· ·Σ\ ^^ =.3⁄4 Α-Λ| Dw will (1 5) Equation (14) can be obtained by solving Kf · Kvj (1 6 ). +13⁄4-tan# When the gain crossover frequency _ is determined, it can be calculated by (I6)

Kf • Kvi. -19- 200928629 Kvj can be calculated by substituting Kf · KVj into (1 3 ), and Kf can be calculated by dividing Kf · Kvj by Kvj. In this way, the closed loop of the first graph can be calculated as the phase margin (5 is stabilized (Kvj, Kf), that is, the stable boundary. When the stable boundary is formed by two curves, the closed loop of Fig. 1 The optimum gain group which is stable and has a large amplitude (KVj, Kf) is the intersection of the above two curves. ❹ On the other hand, when the above-mentioned stable boundary is formed by one curve, the above optimum gain group is When the stability curve is one, the adjustment curve operator 123 sets "the correction amount Kf to zero, and the feedback control gain is the correction amount Kf that can be made to the stable curve. A point that is zero, that is, a line that increases the control gain, and an increase is formed as the above-mentioned adjustment curve. When the above-mentioned stability curve is two, the control gain is obtained by the intersection of two stable curves in the middle of the two stable curves. The optimum 曲线 curve is output as the above-mentioned adjustment curve G. The initial 値 setter 124 is the above-described feedback control of the point on the above adjustment curve and the point at which the control gain is optimally removed from the set enthalpy. The benefit* and the correction amount Kf are output as the control gain initial 値Kin. As described above, the setting of the above-mentioned adjustment curve allows the closed-loop system of Fig. 1 to generate no vibration, and can be quietly and safely (Kvj, Kf). Adjusted to the above optimal gain group. The control gain adjuster 125 is adjusted from the initial point of the control gain 値Kin along the above-mentioned adjustment curve to the above-mentioned stable boundary line, that is, the above-mentioned optimal gain group adjustment -20-200928629 control gain group (KVj, Kf The control gain group (KVj, Kf) is brought close to the above optimal gain group along the above adjustment curve, and the control gain group (KVj, Kf) whose response V is not the oscillation range is the highest 作为 is used as the optimal 値 correction amount. The adjuster 113 calculates the vibration suppression signal vs by multiplying the phase adjustment vibration component by the correction amount Kf. The gain adjustment of the present invention is adjusted from the initial stage of the control gain toward the optimum gain group along the adjustment curve (KVj, Kf), whereby the closed loop system can be quietly and safely adjusted to the above-mentioned maximum gain group in a short time without vibration. In this embodiment, feedback is revealed. The controller 102 is the condition of the speed PI controller, but the same can be applied to the speed P controller, the speed Ι-P controller, the position P speed PI controller, the position P speed Ι-P controller, the position PID control Any controller, etc. In addition, the same can be applied to the above-mentioned torque command chopper, notch filter buffer, disturbance observer, and the situation of no-load time or quantity, and the present invention is A rigid body mechanism or an arbitrary multi-inertial mechanism can be applied in the same manner as the present embodiment. Further, in order to reduce the powers of the equations (13) and (16),

Gn(s) · Gp(s) is approximated as shown in the equation (17), and the above-described stability boundary is calculated in the same manner, and is automatically adjusted (KVj, Kf) based on the stability boundary. 1 s2 + 2 ζ αλω < Α β + ω Ι [ -21 - (17) 200928629 The computer simulation results of the present embodiment are disclosed below. The number of computers used is as follows. J;l· 51X10_4[kg.m2], ω =295Χ2Χ π [rad/s], ζ =0·16, rl rl ω =128.Χ2Χ 7t[rad/s], ζ =0.16, al al ω =1340Χ2Χ ?c[rad/s], ζ =0. 05, γ2 γ2 ω =906Χ2Χ π [rad/s], ζ =0.05, η =10, a2 a2 ds =5_〇Χ2Χ π [rad/s], 1 =6· 28Χ ΙΟ2, ο ι 1 =9. 87Χ1〇\ Kv=6〇X2Xtc[s_1]n 2 J =1. 51X10_4[kg.m2], Ο Κ =KvXJ=0. 0569[N*m*s /rad]% Τ =4/Κν=0. 〇〇85[s]n τ = vj if )=〇. 44X10-3[s], ω =ω =134〇X2XTT[rad/s], ζ =0_7 η r2 η

Trat=0. 637[N*m], T=125X10_6[s]% T=250X10~6[s] t wei, ^ is the pole of the power observer of the Pader approximation used in (10), Li, h is the interference observer φ that is extremely so, φ is the normalized speed proportional control gain, Trat is the rated torque, period, and Tl is the dead time. The nominal moment of inertia Jo is the disturbance moment and the moment of inertia used by the vibration component operator 111. The speed proportional control gain Kvj, the speed control integration time 1 and the torque command filter time constant r f are set as the slave attributes using the normal example control gain κν as described above. The sample used in this computer simulation is as described above in the feedback controller 102 as the speed command filter of the formula (1), and the torque controller 1 〇4 is (: simulation station Make

1/(4ΧΚν , S 〇 is the speed gain, Κν Τ is the control detector 107 and the speed ratio is 1 in the figure, and the (2) type trap-22-200928629 wave filter, the control object 105 is (4 In the equation, the vibration component computing unit 1 is configured to calculate the vibration component from the response V by subtracting the velocity estimation 値V11 calculated by the velocity observer of the equation (7), and the phase adjuster 112 is a bandpass of the formula (5). The filter and the disturbance observer 107 are in the equation (6). Only the discrete time expression of the control period T other than the equation (4) of the control object 105 is expressed, and the torque command Tref is saturated with the rated torque Trat. It is a Bode diagram and a Nyquist diagram showing the open-loop system Q of the computer simulation according to the first embodiment of the present invention. Fig. 3(a) is a Bode diagram of the open-loop transfer function 201 of Fig. 2. In the figure, the solid line is the Bode diagram depicted by the three inertia system samples according to equations (3) and (4), and the one-dotted dashed line is the Bode diagram depicted by the two inertia approximation sample of equation (1 7 ). 3 (a) The figure shows the Bode diagram drawn from the sample of the three inertia system, except for the periphery of the second anti-resonant frequency ω32. It is approximately the same as the Bode diagram depicted by the two inertia approximation samples. In Fig. 3(a), a, b, and c are frequencies with amplitude 1 respectively, and the frequency with minimum phase margin is c. The phase diagram shows that the third (b) is the Nyquist diagram of the above open-loop system. In the third (b) diagram, the solid line is the Nyquist diagram drawn according to the above three inertial system samples, a dotted line It is a Nyquist diagram drawn according to the above two inertia approximation samples, and the broken line is a unit circle. In Fig. 3(b), a, b, and c are points indicating that the Nyquist diagram and the unit circle are intersected, respectively. Corresponding to a, b, and c in Fig. 3(a) respectively, it is known from Fig. 3(b) that point c is the closest to the stability of the unit feedback system of the second quadrant to the 2-23-200928629 graph. When the normalized speed proportional control gain Κν is changed, the phase map of the third (a) graph hardly changes, and only the amplitude map is roughly proportional to the normalized speed proportional control gain Kv variation. Normalized speed proportional control The frequency of the vibration generated when the gain Κν is increased is the frequency of -180 degrees with the above phase map. That is, the phase crossover frequency is approximately the same. 0 The stability boundary operator 1 22 first sets an appropriate 値 for the speed proportional control gain Κν, draws the Bode diagram of the third (a) diagram, and calculates the phase crossover frequency as the vibration. The suppression vibration frequency ω ν suppressed by the suppression unit 1 1 0. Next, the amplitude of the normalization speed proportional control gain Κν which is smaller than 1 is calculated at the phase crossover frequency. The band passing filter time constant r, interference The observer gains h, l2 and the observer gain 1 are set as follows. 11 - (〇y 2 * \ 2 = (〇v / 2 * 1= ίϋ v / 2 * τ = 2 (t) v Based on the above The stabilization boundary is calculated by the stability boundary operator 122. The stability boundary operator 122 calculates an equivalent unit feedback system of the closed loop system including the feedback controller 102 and the control target 1 〇 5 based on the mechanical constant, so that the phase margin of the equivalent unit feedback system can be a positive phase margin. The above setting is used to calculate the above-mentioned stability boundary. The phase margin setting 値 is set such that the above-mentioned uncertainty is allowed to stabilize the closed loop system when the mechanical constant has an uncertain state. 値-24- 200928629 FIG. 4 is a computer simulation of the first embodiment of the present invention. A map of the stability boundary, the adjustment curve, and the initial control gain. In Fig. 4, the solid line indicates the stability boundary, the broken line indicates the adjustment curve, X indicates the initial stage of the control gain, and □ indicates the stable area. This computer simulation uses a phase margin of 5 degrees. As shown in Fig. 4, the stability boundary operator 122 outputs the stable D boundary, and the adjustment curve operator 123 outputs the adjustment curve. The initial threshold setter 14 is the set control gain initial value 値Kin, and the control gain adjuster 125 is From the control gain initial stage 値Kin, the control gain group (Kvj, Kf) is adjusted along the intersection of the above-mentioned adjustment curve toward the above-described stable boundary line, that is, the above-described optimum gain group. When the output mechanical constant of the mechanical constant identifier 121 is not correct, the phase margin 6 is set to be larger than the twist, so that the stable region of FIG. 4 can be narrowed, and the mechanical constant can be allowed to be incorrect. It is possible to adjust the gain quietly and safely without causing vibration in the closed loop system of Fig. 1. In addition, when the control object 105 is an arbitrary multi-inertial system, the closed loop system of Fig. 1 does not include the disturbance observer, and may not be trapped. The presence or absence of the wave filter or the low-pass filter, and in addition, the control rule of the feedback controller 102 can be applied, and the present invention can be applied in the same manner as the embodiment. That is, the feedback controller is the speed P controller, the speed Ι In the case of the -P controller, the position P speed PI controller, the position P speed Ι-P controller, the position PID controller, etc., the present invention can be similarly applied to the present embodiment -25-200928629. As described above, the present invention constitutes a stable region in which the control gain is calculated based on the mechanical constant of the control target. Therefore, vibration of the control target is not generated, the control gain can be adjusted quietly and safely, and automatic and high control performance can be realized. [Industrial Applicability] q The control gain can be adjusted quietly and safely without causing vibration of the motor after the load is connected. Therefore, it can be widely used in semiconductor manufacturing equipment, liquid crystal panel manufacturing equipment, machine tools, and industrial applications. General industrial equipment such as robots. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram showing a schematic configuration of a motor control device according to a first embodiment of the present invention. Fig. 2 is a diagram showing an equivalent unit feedback mechanism of the motor control device according to the first embodiment of the present invention shown in Fig. 1. Fig. 3 is a view showing a Bode diagram and a Nyquist diagram of the open loop system of the computer simulation of the first embodiment of the present invention. Fig. 4 is a view showing the stability boundary, the adjustment curve, and the initial stage of the control gain of the computer simulation of the first embodiment of the present invention. Fig. 5 is a block diagram showing an example of a conventional motor control device. [Description of main component symbols] -26- 200928629 1 〇1: Command generator 1 02: Feedback controller 103: Input signal converter 104: Torque controller 105: Control object 106: Response detector 107: Disturbance observer φ 108 : Frequency response measurement signal generator 1 1 〇: Vibration suppression unit 1 11 : Vibration component calculator 1 1 2 : Phase adjuster 1 13 : Correction amount adjuster 120 : Control gain adjustment unit 1 2 1 : Mechanical constant recognizer 122 '·stability boundary operator ❹ 123 : adjustment curve operator 124 : initial 値 setter 125 : control gain adjuster 201 : open loop transfer function 501 : target for gain automatic adjustment 値 command unit 502 : position proportional control gain 5 03 : Speed proportional control gain 504 : Torque control unit 5 05 : Encoder -27 - 200928629 5 0 6 : Motor 507 : Ball screw mechanism 5 08 : Position deviation evaluation function calculation unit 509 : Position deviation evaluation function Memory unit 510 : Evaluation function Minimum gain determination unit 511: Torque command evaluation function calculation unit 512: Torque command evaluation function storage unit 5 1 3 : Torque command determination unit -28-

Claims (1)

200928629 X. Patent Application No. 1 - A motor control device comprising: a feedback controller that synchronizes a command (vr) with a response (v) of a control object to perform a control calculation, and calculates a torque command (τ«) before interference correction; Interference observer for calculating the interference estimation 値(Td) based on the 扭矩 immediate torque command (Tref) and the above response (V) 0 after subtracting the interference estimation 値(Td) from the disturbance correction torque command (Τα); In response to the measurement signal (Fr) and the torque command (Tref), the frequency response measurement signal (Fr) is output as a current command (I) during frequency response measurement, and the torque command (Tref) is used as a current during gain adjustment. An input signal converter for outputting an instruction (Ir); a torque controller for driving the motor flow motor current (Im) of the control target according to the current command (Ir); and according to the response (V) and the torque command (Tref) The vibration suppression unit that calculates the vibration suppression Q signal (Vs) is configured to add the vibration suppression signal (Vs) to the command (Vr), which is characterized by The control gain adjustment unit that calculates the feedback controller control gain (KVj) and the correction amount (Kf) is calculated based on the response (V). 2. The motor control device according to claim 1, wherein the frequency response measurement signal (Fr) is a frequency sine wave signal. 3. The motor control device according to claim 1, wherein the frequency response measurement signal (Fr) is a complex frequency sine wave. The motor control device according to the first aspect of the invention, wherein the control gain adjustment unit is configured to: control the machine to be controlled based on the response (V) during frequency response measurement a mechanical constant identifier for identifying a constant; calculating, based on the mechanical constant, a stability of a stable boundary line of the feedback control gain including the feedback control of the feedback controller and the control target, and a stable region outline of the correction amount (Kf) The boundary line operator calculates an adjustment curve capable of adjusting the feedback control gain and the correction amount (Kf) based on the stability boundary to improve the adjustment curve of the closed-loop responsiveness and the interference suppression characteristic while maintaining the stability of the control target. An arithmetic unit that calculates the feedback gain (KVj) of the control gain adjustment and the initial value of the correction amount (Kf), that is, the initial setting of the control gain initial value (Kin) based on the adjustment curve, and the adjustment curve according to the adjustment curve The above control gain initial 値 (Kin) adjusts the above feedback control Gain (Kvj) and the correction amount (Kf of) control gain adjuster. The motor control device according to claim 4, wherein the mechanical constant identifier is a frequency response for calculating the control target, and the mechanical constant is calculated based on the frequency response. 6. The motor control device according to claim 4, wherein the stability limit line operator calculates an equivalent unit feedback system of the closed loop system based on the mechanical constant, and causes a phase margin of the equivalent unit feedback system The stability margin can be calculated by setting the phase margin of a positive number. The motor control device according to the sixth aspect of the invention, wherein the phase margin setting 设定 is set such that the uncertainty can be allowed when the mechanical constant is uncertain. The motor control device according to the fourth aspect of the invention, wherein the adjustment curve operator has "the correction amount (Kf) is zero when the stability curve is one. When the feedback control gain (Kvj) is 0, a point at which the correction amount (Kf) of the stability curve is zero, that is, a line where the control gain is optimally increased, is formed as the adjustment curve output, and when the stability curve is two 'The curve of the control gain optimum 成为 which is the intersection of the two stable curves in the middle of the two stable curves is output as the above adjustment curve. The motor control device according to claim 8, wherein the initial setting device is the feedback control at a point on the adjustment curve and from a point at which the control gain is optimally set to a predetermined amount. The gain ❹ (Kvj ) and the above-described correction amount (Kf ) are output as the control gain initial 値 (Kin). 10. The motor control device according to claim 9, wherein the control gain adjuster increases the feedback control gain from the initial control gain (Kin) toward the control gain along the adjustment curve. (Kvj) and the above correction amount (Kf) are then output. -31 -