JP4196400B2 - Motor control device and control method - Google Patents

Description

【００２８】
ここで、ｍ_iはリンクｉの質量、Ｌ_iは軸から重心までの距離、ａ_iはリンク長さである。また、速度に関する干渉力の項は速度に依存しており、慣性項に比較して変化は緩やかであり、速度制御ループで補償されるため、無視して考えられる。その場合、式（１）と（２）は次のように表せる。
【００２９】
【数２】

【００３０】
よって、式（３）と（４）は以下のように表現することができる。
【００３１】
【数３】

【００３２】
ここで、
【００３３】
【数４】

【００３４】
また、Ｉ_iは重心周りの慣性モーメント、Ｎ_iはｉ軸モータの減速機の減速比、Ｉ_imはモータのモータロータのイナーシャ（等価慣性モーメント）である。
ｍ_i、Ｌ_i、ａ_i、Ｎ_iは既知であり、主慣性であるＪ₁₁とＪ₂₂は、負荷イナーシャＪＬとして既に求められているため、特定の姿勢θ₁とθ₂での重心周りの慣性モーメントＩ_iを式（６）〜（８）から求め、相互慣性であるＪ₁₂とＪ₂₁を求めることができる。
【００３５】
式（５）は、加減速時に２軸目が１軸目に作用する干渉トルクτ₁₂と１軸目が２軸目に作用する干渉トルクτ₂₁が、以下のように表せることを意味する。
【００３６】
【数５】

【００３７】
よって、図８に示すように、求められた相互慣性のパラメータＪ₁₂（＝Ｊ₂₁）を元に、実機部10に相互干渉部35を配置する。
【００３８】
（２）相互慣性モデルのパラメータＪ₁₂の調整
上位コントローラから、通常運転時の位置（角度）指令を、前記実機部10の前記モータ制御部11a、11b及び前記シミュレーション部20の前記制御モデル部21a、21bに入力する。この指令は、図７(b)に示すように１軸目を・・θ₁で、２軸目を・・θ₂で動作させることで、２軸目に干渉力τ₂₁を与え、実機部とシミュレーション部の挙動（軌跡ズレ）を比較する。
実機部の相互干渉補償部35の相互慣性Ｊ_r12（＝Ｊ_r21）と、実際に実機部の機構部に作用している相互慣性Ｊ_s12（＝Ｊ_s21）が一致していない場合には、実機部とシミュレーション部に軌跡ズレが発生する。ここでは、実機部とシミュレーション部の軌跡ズレを定量的に判断することで、相互干渉補償部35の相互慣性パラメータを調整する。
【００３９】
実機部とシミュレーション部の軌跡の比較をするために、順変換部34では、前記モータ制御部11の位置検出器14a、14bの位置ＦＢと、前記制御モデル部21の位置検出器モデル24a、24bの位置ＦＢをそれぞれ順変換し、直交座標系における実機部とシミュレーション部のロボット先端の位置ＦＢ（状態量）を求める。前記状態比較部31では、２つの前記位置ＦＢを差分して状態量比較値を求め、評価量演算部32で予め決められたしきい値以内であるかどうかを判断する。
しきい値以内であれば、制御パラメータ（相互慣性）の調整を終了する。これで実際に作用している相互干渉と実機部の相互干渉補償部35が近似値になったことになる。しきい値以上であれば、以下に示すように制御パラメータ調整部33で相互干渉補償部35の制御パラメータＪ₁₂を変更し、しきい値以内になるまで繰り返す。
【００４０】
相互慣性のパラメータＪ₁₂の変更方法は、図７(b)に示すように実機部の軌跡とシミュレーション部の軌跡のズレ量（ΔＸ，ΔＹ）から、相互慣性のパラメータＪ₁₂を変更する大きさと正負方向について判断する。これには、直交座標系における軌跡のズレ量を、ヤコビアンと呼ばれる関節座標系と作業座標系の微小変位関係式の逆行列を用いて、関節座標系のズレ量に変換する。２軸のヤコビアンJacobianは以下の式で表すことができる。
【００４１】
【数６】

【００４２】
ここで、θ₁₂＝θ₁＋θ₂
また、その逆行列Jacobean^-1は以下の式で表すことができる。
【００４３】
【数７】

【００４４】
よって、直交座標系における軌跡のズレ量（ΔＸ，ΔＹ）と関節座標系における角度誤差（Δθ₁，Δθ₂）は以下のように表すことができる。
【００４５】
【数８】

【００４６】
また、関節座標系の剛性から、角度誤差とトルクの関係は以下の式で表すことができる。
【００４７】
【数９】

【００４８】
ここで、Ｋ＝Ｋｐ・Ｋｖ・Ｊ
よって、相互慣性のパラメータＪ₁₂は、式（１３）と以下の式を解くことで得ることができる。
【００４９】
【数１０】

【００５０】
実際の作業で使用する位置指令を用いて以上の作業を繰り返すことで、多軸ロボットの相互干渉による干渉力補償の制御パラメータを調整することができる。
【００５１】
【発明の効果】
以上述べたように、請求項１のモータの制御装置によれば、実機部及びシミュレーション部を並列に配置し、入力された同一の指令に応じて機構部及び機構部モデルの状態量を比較し、予め決められたしきい値以内に状態量入るまで、同一指令による動作の比較から実機部の制御パラメータ値の変更を繰り返すことにより、重力や動摩擦等の影響を受けずに、動作領域の制限などの制約を受けずに、制御部の制御パラメータを自動的に調整することができる。
請求項２記載のモータの制御装置によれば、複数のモータ制御の状態量で調整動作を実施することにより、複数軸の場合の軸間に生じる制御パラメータも調整することができる。また、複数の前記モータの相互干渉を補償し、相互干渉のパラメータを調整することにより、相互干渉の制約を受けずに、制御部の制御パラメータを自動的に調整することができる。
請求項３記載のモータの制御装置によれば、前記モータ制御部のゲイン調整は前記モータ制御部の速度、加速度の振動特性が前記シミュレーション部の同信号に比較して一定レベル以下になるように調整することにより、状況を的確に判断して運転条件に依らず、正確にモデルパラメータを同定できる。
請求項４記載のモータの制御装置によれば、前記振動特性の評価はＦＦＴなどの振動周波数成分の大きさを評価する手法を用いることにより、問題となる振動成分を的確に判断して、正確に制御パラメータの調整を行うことができる。
請求項５記載のモータの制御装置によれば、振動特性の評価を実施後、前記モータ制御部の制御構造を変化させることにより、実機及びモデルの機構構成に合った制御系を構築でき、制御対象をより的確に制御できる。
請求項６記載のモータの制御装置によれば、前記評価量演算部の判断により、前記調整動作の実行を選択することにより、自動的にパラメータの更新ができる。
請求項７記載のモータの制御装置によれば、前記モータ制御部の制御パラメータを位置ループゲイン又は速度ループゲインとすることにより、高精度の位置決めや速度応答が可能となる。
請求項８記載のモータの制御方法によれば、実機部及びシミュレーション部を並列に配置し、入力された同一の指令に応じて機構部及び機構部モデルの状態量を比較し、予め決められたしきい値以内に状態量入るまで、同一指令による動作から実機部の制御パラメータ値の変更を繰り返すことにより、重力や動摩擦等の影響を受けずに、機構部の周りに複数の周辺機器がある場合でも動作領域の制限などの制約を受けずに、制御部の制御パラメータを簡単に調整することができる。
請求項９記載のモータの制御装置によれば、複数のモータ制御の状態量で調整動作を実施することにより、複数軸の場合の軸間に生じる制御パラメータも調整することができる。また、複数の前記モータの相互干渉を補償し、相互干渉のパラメータを調整することにより、相互干渉の制約を受けずに、制御部の制御パラメータを自動的に調整することができる。
【図面の簡単な説明】
【図１】本発明の第１の基本構成図
【図２】本発明の第１の具体的実施例を表すブロック図
【図３】本発明の第１の具体的実施例における応答波形１
【図４】本発明の第１の具体的実施例における応答波形２
【図５】本発明の第２の具体的実施例における応答波形
【図６】本発明の第３の具体的実施例を表すブロック図
【図７】本発明の第４の具体的実施例を表すモデル
【図８】本発明の第４の具体的実施例を表すブロック図
【図９】従来の制御方式を示す図
【符号の説明】
１０：実機部
１１：モータ制御部
１２：機構部
１３：モータ
１４：位置検出器
１５：外力補償部
２０：シミュレーション部
２１：制御モデル部
２２：機構モデル部
２３：モータモデル
２４：位置検出器モデル
２５：外力補償部モデル
３１：状態量比較部
３２：評価量演算部
３３：制御パラメータ調整部
３４：パラメータ同定部
３５：相互干渉補償部
１１１：位置速度ループ
１１２：アンプ
１２１：２次側機構部
１２２：減速機
２１１：位置速度制御ループモデル
２１２：アンプモデル
２２１：２次側機構部モデル
２２２：減速機モデル
３１１：振動成分抽出部[0001]
[Industrial application fields]
The present invention relates to a motor control device, and more particularly to a motor control device and a control method for adjusting motor control parameters.
[0002]
[Prior art]
A block diagram of a motor control system applied to the motor control device is shown in FIG. Since the characteristics of the motor control system directly affect the overall operation including the mechanism system and the motor, adjustment of control parameters of the control device is extremely important when high trajectory accuracy and speed responsiveness are required.
A method for adjusting control parameters, in particular, control parameters such as position gain and speed gain will be described. In a motor control system, it is necessary to determine a loop gain as large as possible while taking care not to generate oscillation from the inner loop in order to keep good steady-state characteristics, quick response and rigidity. Conventionally, an appropriate gain has been set by an operator so as not to generate vibration or abnormal noise from the motor. This is because when a specific gain is input in advance and the motor performs an operation close to actual use, if a vibration or abnormal noise occurs, the gain is reduced and a response delay in position or speed occurs. Is a method of adjusting the gain so that the desired motor performance can be obtained by taking steps such as changing the gain greatly.
[0003]
Further, as a conventional example 1 of a device for automatically adjusting the control gain of a conventional motor, an integration of a test pattern generator for generating a pattern for changing the rotational speed of the motor (including a ramp-shaped speed command) and a speed control calculator A switch for shutting off the output to the tester, a speed loop gain multiplier capable of switching to the test gain, and a gain adjuster. In this control device, a load inertia is adjusted by a gain adjuster that changes the loop gain of the control loop of the motor in accordance with the output difference from the test pattern of the ramp-shaped speed command actually driven and the speed detection means for detecting the rotational speed of the motor. , And auto-tuning that sets a loop gain more suitable for this load inertia becomes possible (see Patent Document 1).
Further, as a conventional example 2, in a control system that performs proportional control of a control object by adding a value obtained by multiplying the speed deviation to the gain of the control loop to the control object, from the oscillation region toward the stable region with respect to an increase in the speed deviation. A gain adjustment system having a characteristic of decreasing the gain is formed, and a gain that converges to the oscillation limit is obtained (see Patent Document 2).
[0004]
[Patent Document 1]
JP-A-6-70566
[Patent Document 2]
JP-A-9-106303
[0005]
[Problems to be solved by the invention]
However, in the conventional method of adjusting the gain of the control parameter, it is necessary to rely on the experience of a skilled person, and depending on the operator to be tuned, not only does it take a lot of time to obtain the optimum gain, There is a problem that the motor control system may oscillate. In addition, in order for the operator to confirm the occurrence of vibration, it has been necessary to measure information on the position and speed of the motor using a separate measuring instrument.
Further, the conventional example 1 has a problem that a special adjustment operation is required when adjusting the parameters, and the parameters cannot be adjusted during normal operation. When the load on the mechanism section is frequently changed, it is necessary to stop the normal operation each time, secure an operation range in which parameters can be adjusted, and execute a special adjustment operation. Even if adjustment is performed using a dedicated operation command, there is a risk that if the operation command is used in actual work, the control system may oscillate due to a difference in conditions. Furthermore, in terms of accuracy, when friction acts on the drive system including the motor, there is a problem that the parameter adjustment cannot be made accurate because the position deviation and the speed detection value are affected.
[0006]
Similarly, the conventional method 2 also has a problem that the parameter adjustment cannot be made accurately because the speed deviation is affected when gravity or friction acts on the motor in terms of accuracy. There is also a problem that it only supports adjustment of the speed loop gain of the control parameter.
Therefore, the present invention aims to adjust control parameters with high accuracy in a short time even during normal operation and under various conditions on the load side without requiring any special operation for adjustment.
[0007]
[Means for Solving the Problems]
  In the present invention, in order to solve the above problems, the following configuration is adopted.
  The motor control device according to claim 1 drives the motor based on a command from the motor control unit and is connected to the motor.MachineIn a motor control apparatus having a structure part, a simulation unit having a motor model simulating the motor, a mechanism model part simulating the mechanism part, a control model part simulating the motor control part, and the motor control The same state quantity of the actual machine part and the simulation part consisting of the motor part, the motor and the mechanism partAny one of motor position, motor speed, or motor accelerationCompareIt is a vibration component of the state quantity extracted using FFTA state quantity comparison unit that creates a state quantity comparison value is compared with the state quantity comparison value and a preset threshold value. When the state quantity comparison value is equal to or greater than the threshold, the control of the motor control unit is performed. An evaluation amount calculation unit that determines that the parameter is not optimal and determines that the control parameter is optimal when the state quantity comparison value is within the threshold; and the control parameter is not optimal in the evaluation amount calculation unit And a control parameter adjustment unit that adjusts the control parameter so that the state quantity comparison value falls within the threshold value when judged.
[0008]
  The motor control device according to claim 2 is a motor control device that has a mechanism unit that drives the motor based on a command from the motor control unit and is connected to the motor, and includes a plurality of motors connected to each other. In a motor control device for driving an axis robot, an actual machine unit comprising the motor control unit, the motor, and the mechanism unit, and further including a mutual interference compensation unit having a mutual inertia parameter for compensating for mutual interference between the axes. A simulation unit having a motor model simulating the motor, a mechanism model unit simulating the mechanism unit, and a control model unit simulating the motor control unit for each axis, the actual unit, and the simulation unit A state quantity comparison unit that compares the same state quantity and creates a state quantity comparison value, compares the state quantity comparison value with a preset threshold value, and compares the state quantity ratio. An evaluation amount calculator that determines that the mutual inertia parameter is not optimal when a value is equal to or greater than the threshold, and that determines that the mutual inertia parameter is optimal when the state quantity comparison value is within the threshold; A control parameter adjustment unit that adjusts the mutual inertia parameter so that the state quantity comparison value falls within the threshold when the evaluation amount calculation unit determines that the mutual inertia parameter is not optimal. Is.
  According to a third aspect of the present invention, there is provided the motor control device according to the second aspect, wherein the state quantity is any one of a motor position, a motor speed, and a motor acceleration.
  According to a fourth aspect of the present invention, there is provided the motor control apparatus, wherein the state quantity comparison unit according to the second aspect uses the vibration component of the state quantity extracted using FFT as the state quantity comparison value. .
  According to a fifth aspect of the present invention, there is provided the motor control apparatus according to the first aspect, wherein, even if the control parameter adjustment unit adjusts the control parameter, the state quantity comparison value is equal to or greater than the threshold. The control device changes the control structure of the motor control unit, the model structure of the mechanism model unit, and the control structure of the simulation unit.
  According to a sixth aspect of the present invention, in the motor control device according to the first or second aspect, the evaluation amount calculation unit monitors the state quantity comparison value, and determines the state quantity comparison value according to the magnitude of the state quantity comparison value. Adjustment of the control parameter or the mutual inertia parameter is automatically started.
  According to a seventh aspect of the present invention, there is provided the motor control apparatus according to the first aspect, wherein the control parameter is a position loop gain or a speed loop gain.
[0009]
  The motor control device according to claim 8 includes a mechanism unit that drives the motor based on a command from the motor control unit and is connected to the motor, and includes the motor control unit, the motor, and the mechanism unit. In the control method of a motor control device including a real machine unit, the motor control device further includes a simulation unit having a model simulating the real machine unit, and inputs the same command to the real machine unit and the simulation unit And the same state quantity of the actual machine part and the simulation partAny one of motor position, motor speed, or motor accelerationCompare, The vibration component of the state quantity extracted using FFTA state quantity comparison value is created, the state quantity comparison value is compared with a preset threshold value, and the control parameter of the motor control unit is not optimal when the state quantity comparison value is equal to or greater than the threshold. Determining that the control parameter is optimal when the state quantity comparison value is within the threshold, and determining that the control parameter is not optimal when the control parameter is not optimal. The process is performed by adjusting the control parameter so as to enter.
  The motor control device according to claim 9 includes a mechanism unit that drives the motor based on a command from the motor control unit and is connected to the motor, and includes the motor control unit, the motor, and the mechanism unit. In a control method of a motor control apparatus that drives a multi-axis robot having a real machine section and connected to a plurality of the motors, the motor control apparatus further compensates for mutual interference between axes in the real machine section. A mutual interference compensation unit having a mutual inertia parameter, and a simulation unit having a model simulating the real machine unit for each axis, and inputting the same command to the real machine unit and the simulation unit. A state quantity comparison value is created by comparing the same state quantity with the simulation unit, the state quantity comparison value is compared with a preset threshold value, and the state quantity comparison value is the threshold value. In the above case, it is determined that the mutual inertia parameter is not optimal, and when the state quantity comparison value is within the threshold, it is determined that the mutual inertia parameter is optimal, and it is determined that the mutual inertia parameter is not optimal. In this case, the processing is performed by adjusting the mutual inertia parameter so that the state quantity comparison value falls within the threshold value.
[0010]
DETAILED DESCRIPTION OF THE INVENTION
The basic configuration of the present invention will be described with reference to FIG. Here, for the sake of simplification of explanation, it is assumed that one motor is driven by the present control device, and the mechanism connected to the motor is a simple cylindrical load via a speed reducer.
In FIG. 1, the actual machine unit 10 includes a mechanism unit 12, a motor 13 connected to the mechanism unit 12, and a motor control unit 11 that controls the motor 13. The simulation unit 20 includes a motor model 23 that simulates the dynamic system of the motor 13, a mechanism model unit 22 that simulates the dynamic system of the mechanism unit 12, and a control model unit 21 that simulates the motor control unit 11. It is configured.
In the device that controls the motor of the control device, the actual machine unit 10 and the simulation unit 20 are arranged in parallel, and each receives the same command from the host controller. The actual machine unit 10 and the simulation unit 20 output torque commands for controlling the mechanism unit 12 and the mechanism unit model 22 to the motor 13 and the motor model 23, respectively, according to the input commands. Here, the same command input from the host controller is an operation command used in actual work, and a pattern stored in advance may be reproduced or used, or a method of automatically creating from CAD data or the like. But it ’s okay. Using a pre-stored pattern, for example, has an advantage that there is no collision with the peripheral device during the identification operation because interference with the peripheral device is avoided at the time of pattern creation.
[0011]
In the state quantity comparison unit 31, the same state quantity (for example, position deviation and speed feedback) of the motor control unit 11 and the control model unit 22 in the actual machine unit 10 and the simulation unit 20 or the mechanism unit 12 is used. Are compared with the same state quantity (for example, position feedback) of the mechanism model unit 22 to create a state quantity comparison value.
Whether the control parameter of the motor control unit 11 of the actual machine unit 10 is optimally set based on whether or not the state quantity comparison value created by the state quantity comparison unit 31 is within a threshold value. Judge whether. Here, when the parameters of the real machine and the model are completely the same, that is, when the real machine and the simulation exhibit exactly the same behavior, the determination criterion is that the state quantity comparison value is ideally 0.
[0012]
Here, when comparing with the threshold value, it is possible to select whether to use the instantaneous value of the state quantity comparison value or the accumulated value of the state quantity comparison value. Further, in the case of a mechanical device such as a robot to which a plurality of motors are connected, a value obtained by converting the state quantity comparison value into an orthogonal coordinate value can be used as the state quantity of the hand effector portion of the robot.
If the state quantity comparison value is equal to or greater than the threshold value, the control parameter adjustment unit 33 changes the control parameter values of the motor control unit 11 and the control model unit 21 in conjunction with each other, and the state quantity comparison value becomes the threshold value. The control parameter value is changed from the operation with the same command until it falls within the value. If the state quantity comparison value is within the threshold value, it is determined that the control parameters of the actual machine and the simulation are approximate values, and the adjustment operation is terminated. Here, as a method for changing the control parameter value, a method is used in which the correction direction is determined by the sign of the state quantity comparison value, and the correction amount is determined by the magnitude of the state quantity comparison value with respect to the threshold value.
[0013]
The operator can select whether to perform the adjustment operation during the normal work or to interrupt the normal work and execute the special adjustment operation through the control device. Further, during normal operation, the state quantity comparison value is monitored successively in the evaluation quantity calculation unit 32, and the necessity / unnecessity of adjustment is determined by the state quantity comparison value increasing, and the adjustment is automatically performed. The operation may be executed.
[0014]
Hereinafter, a first specific embodiment of the present invention will be described with reference to FIG.
The motor control unit 11 includes a position / speed control loop 111 and an amplifier 112. By inputting a position (angle) command for normal operation and outputting a torque command to the amplifier 112, the motor 12 and The mechanism unit 13 is driven. In the position / speed control loop 111, for example, position proportional-speed proportional integral control shown in FIG. 9 is used. Similarly, the control model unit 21 also includes a position / speed control loop model 211 and an amplifier model 212. By inputting a position (angle) command during normal operation and outputting a torque command to the amplifier model 212, A model structure for driving the mechanism model unit 22 is adopted.
[0015]
Furthermore, when gravity acts on the mechanism assumed in the basic configuration, when the reduction gear friction is large, or when there is a high demand for adjustment accuracy, the difference between the actual unit and the simulation unit is eliminated. It is also possible to perform compensation. The motor control unit 11 incorporates an external force compensation unit 15 for gravity, static friction and dynamic friction of the mechanism unit 12, and calculates gravity torque and friction torque according to position feedback (hereinafter referred to as FB) to obtain torque. By adding to the command, the effect of external force is compensated. Alternatively, the control model unit 21 incorporates the gravity model of the motor model 23 and the external force compensation unit model 25 of the static friction and dynamic friction models, and calculates the gravitational torque and friction torque according to the command and adds them to the torque command. By doing so, you may simulate the influence of external force. Whether to enable the external force compensator 15 or the external force compensator model 25 is selected as necessary.
Similarly, when the control target is a robot having a plurality of links, the posture change is caused by a normal operation command, so the load inertia of each axis varies in real time. Therefore, by having a model of the mechanism unit in advance, solving the dynamics calculation in real time, and obtaining the load inertia for each posture, the influence of the posture change is compensated when adjusting the control parameter.
Hereinafter, a method for adjusting the speed loop gain Kv and the position loop gain Kp of the motor control unit 11 will be described. Here, the adjustment operation is executed using a command during normal operation, but a special command for adjustment operation may be used.
[0016]
(1) Preparation for adjustment
In the motor control unit 11 of the actual machine unit 10, Kp is used as an initial value of the position loop gain and speed loop gain before adjustment.₀, Kv₀Set. The control model unit 21 of the simulation unit 20 also has the same Kp as the actual unit 10 as initial values of the position loop gain and the velocity loop gain.₀, Kv₀Set.
A position command during work shown in FIG. 3A is input from the host controller to the motor control unit 11 of the actual machine unit 10 and the control model unit 21 of the simulation unit 20. This command is, for example, a movement command from one point during work to another point.
[0017]
(2) Adjustment of speed loop gain Kv
For the adjustment, the constant velocity section (steady state) of the position command is used, and the differential value of the position FB detected by the position detector 14 and the position detector model 24 connected to the motor 13 and the motor model 23 for this command ( Speed FB) is a state quantity. The speed command and speed FB of the actual machine unit and the simulation unit have waveforms shown in FIG.
Here, when a response due to the frequency component of the command occurs, a response of the same frequency occurs at both the speed FB of the actual machine unit and the speed FB of the simulation unit, and the influence of the frequency component caused by the command can be removed. .
The state machine comparison unit 31 compares the actual machine and the simulation speed FB to create a state quantity comparison value (difference in speed FB). It is determined whether or not the obtained state quantity comparison value is within a threshold value determined in advance by the evaluation quantity calculation unit 32. Here, since a high-frequency vibration component is used as a determination criterion, instantaneous values are used for the state quantity comparison.
[0018]
Speed loop gain Kv of actual machine₀Is higher than the appropriate value, the control system becomes oscillating and the vibration of the high frequency component is added to the speed FB, so the value immediately before the high frequency component appears is the optimum value. Therefore, when the high frequency component exceeds the threshold value, the speed loop gain Kv of the motor control unit 11 and the control model unit 21₀Is a small value Kv₁Repeat the adjustment with each setting again. During this adjustment, the value immediately after the value is within the threshold value from the state that is equal to or more than the threshold value is the optimum value Kv.₂To finish the adjustment.
Conversely, when there is no high-frequency component, the state quantity comparison value is equal to or less than the threshold value, but it cannot be determined whether it is an optimal value. Therefore, gradually the speed loop gain Kv of the actual machine part₀Is a large value Kv₁Repeat the adjustment until high frequency components appear. The value immediately before the threshold value is exceeded from the state within the threshold value during the adjustment is the optimum value Kv.₂To finish the adjustment.
By the above procedure, the optimum value Kv of the speed loop gain₂Can be requested.
[0019]
(3) Adjustment of position loop gain Kp
After adjusting the speed loop gain Kv, the position loop gain Kp is adjusted. For adjustment, the position command constant velocity section (steady state) shown in Fig. 4 (a) is used and detected by the position detector 14 and the position detector model 24 connected to the motor 13 and motor model 23 corresponding to this command. The differential value (velocity FB) of the position FB is used as the state quantity. The speed command and speed FB of the actual machine unit and the simulation unit have waveforms shown in FIGS. 4 (b) and 4 (c).
The state machine comparison unit 31 compares the actual machine and the simulation speed FB to create a state quantity comparison value (deviation of the speed FB). It is determined whether or not the obtained state quantity comparison value is within a threshold value determined in advance by the evaluation quantity computing unit 32 and whether a high frequency component is present. As the state quantity comparison value, an average comparison can be performed by using an accumulated value for every predetermined time.
[0020]
Actual part position loop gain Kp₀The optimum value of is the value immediately before the appearance of the high frequency component. Actual part position loop gain Kp₀4 is higher than the optimum value, the control system oscillates as shown in FIG. 4 (b), and the vibration of the high frequency component is added to the deviation of the speed FB, and the state quantity comparison value as shown in FIG. 4 (d). Becomes larger than the threshold value. In this case, the position loop gain Kp of the motor control unit 11 and the control model unit 21₀Is a small value Kp₁Set to and repeat the adjustment. During this adjustment, the value immediately after the value is within the threshold from the state that is equal to or greater than the threshold is set to the optimum value Kp.₂To finish the adjustment.
Conversely, the position loop gain Kp of the actual unit₀4 is smaller than the optimum value, as shown in FIG. 4C, high-frequency component vibration does not occur, but it cannot be determined whether it is the optimum value. Further, since the deviation of the speed FB becomes large, the state quantity comparison value becomes larger than the threshold value as shown in FIG. In this case, gradually the position loop gain Kp of the actual machine part₀Is a large value Kp₁And repeat the adjustment until the vibration of the high frequency component is applied. The value immediately before the threshold value is exceeded from the state within the threshold value is the optimum value Kp.₂To finish the adjustment.
[0021]
Implementation of these series of adjustment operations may be performed by a method in which the operator intentionally executes the operation of the control device when the worker changes the load of the mechanism unit, or in advance in the evaluation amount calculation unit 32. A method is conceivable in which the state parameter comparison value (deviation of the speed FB between the actual machine unit and the simulation unit, etc.) is monitored to automatically detect the fluctuation of the load inertia and start the control parameter adjustment operation. Therefore, in the above embodiment, the deviation of the speed FB during normal operation may be monitored, and if there is a difference in the state quantity between the actual machine unit and the simulation unit, the gain adjustment operation may be automatically started. .
[0022]
The second specific example of the present invention will be described below.
This embodiment differs from the first embodiment in the process of determining a high frequency component when adjusting the position loop gain and the velocity loop gain. In the first embodiment, the extraction of the vibration component is obtained by directly measuring the time change of the waveform. In this embodiment, however, the state quantity comparison unit 31 uses an FFT (Fast Fourier Transform), so The vibration component of the position FB and the speed FB of the simulation unit can be determined more easily.
The vibration determination at the speed FB will be described as an example. By performing an FFT analysis in the vibration component extraction unit 311 in the state comparison unit 31, a high frequency component is easily generated as shown in FIG. 5 (c) from the velocity FB in FIGS. 5 (a) and 5 (b). Can be identified.
Thereafter, according to the procedure described in the first specific embodiment, the control parameter adjustment unit 33 controls the speed loop gain Kv and the frequency and power so that the frequency and power of the actual unit and the simulation unit substantially match within the threshold value. The position loop gain Kp is changed and repeated so that the state quantity comparison value falls within the threshold value.
[0023]
The third specific example of the present invention will be described below.
In the present embodiment, after adjusting the control parameters of the motor control unit obtained in the first specific embodiment and evaluating the vibration characteristics, the motor control unit 11 and the simulation of the actual machine unit 10 are performed as necessary. The control structure of the mechanism model unit 22 of the unit 20 is changed to improve the adjustment accuracy.
When vibration does not become smaller than the threshold value in the vibration characteristic evaluation, it is determined that the order of the model to be controlled does not match, and the order of the mechanism model unit 22 of the simulation unit 20 is increased as shown in FIG. Change direction.
[0024]
Here, the control configuration is changed from a rigid body model (one inertia system) to a two inertia system model. However, when controlling the control object of the two-inertia system, the spring element that connects the two inertias causes vibrations, so it is necessary to subtract the reaction force from the spring from the torque command of the motor control system. Therefore, in the motor control unit 11 of the actual machine unit 10, a two-inertia model observer is arranged, and the estimated twist angle θ_SReaction force τ_RTo a structure that subtracts from the torque command. Further, in the simulation unit 20, the twist angle θ input to the secondary side mechanism unit model 221._SSReaction force τ_SIs subtracted from the torque command.
As described above, by changing the control configuration of the motor control unit 11 and the mechanism model unit 22 according to the adjustment result of the control parameter, vibration on the load side can be suppressed and highly accurate control can be performed.
[0025]
The fourth specific example of the present invention will be described below. Here, a description will be given of a method for adjusting the control parameter for interference force compensation based on the mutual inertia of the robot based on the locus deviation information between the actual machine unit and the simulation unit.
In the present embodiment, a control system parameter adjustment method for compensating for mutual inertia will be described for a horizontal articulated robot to which a plurality of motors are connected. In order to simplify the description, as shown in FIG. 7 (a), the mechanism unit to be controlled is a biaxial horizontal articulated robot. Here, it is assumed that the main inertia of each axis (load inertia JL of each axis) has already been obtained, and a method of adjusting the compensation parameter for mutual inertia that affects other axes will be described.
In the present embodiment, as shown in FIG. 8, a mutual interference compensation unit 35 is provided in the actual machine unit. Since the mutual interference does not act on the simulation part which is an ideal space, if the mutual interference acting on the actual machine part can be completely compensated by the mutual interference compensation part 35, the behavior becomes the same as the simulation part. By utilizing this, the position FB value in the Cartesian coordinate system of the robot tip of the real machine part and the simulation part is compared, and the control parameter for mutual interference compensation of the real machine part is adjusted by reducing the position FB deviation. This will be described in detail below.
[0026]
(1) Setting of mutual interference part
The equation of motion of the two-axis horizontal articulated robot is expressed as follows.
[0027]
[Expression 1]

[0028]
Where m_iIs the mass of link i, L_iIs the distance from the axis to the center of gravity, a_iIs the link length. Further, the term of the interference force related to the speed depends on the speed, and the change is more gradual than the inertia term and is compensated by the speed control loop, so it can be ignored. In that case, equations (1) and (2) can be expressed as follows.
[0029]
[Expression 2]

[0030]
Therefore, equations (3) and (4) can be expressed as follows:
[0031]
[Equation 3]

[0032]
here,
[0033]
[Expression 4]

[0034]
I_iIs the moment of inertia around the center of gravity, N_iIs the reduction ratio of the reducer of the i-axis motor, I_imIs the inertia (equivalent moment of inertia) of the motor rotor of the motor.
m_i, L_i, A_i, N_iIs known and is the main inertia J₁₁And J_{twenty two}Has already been determined as the load inertia JL,₁And θ₂Moment of inertia I around the center of gravity_iIs obtained from the equations (6) to (8), and J is the mutual inertia.₁₂And J_{twenty one}Can be requested.
[0035]
Equation (5) is the interference torque τ that the second axis acts on the first axis during acceleration / deceleration₁₂And the interference torque τ that the first axis acts on the second axis_{twenty one}Means that it can be expressed as follows.
[0036]
[Equation 5]

[0037]
Therefore, as shown in FIG. 8, the obtained mutual inertia parameter J₁₂(= J_{twenty one}), The mutual interference unit 35 is arranged in the actual machine unit 10.
[0038]
(2) Parameter J of mutual inertia model₁₂Adjustment of
A position (angle) command during normal operation is input from the host controller to the motor control units 11a and 11b of the actual machine unit 10 and the control model units 21a and 21b of the simulation unit 20. As shown in Fig. 7 (b), this command gives the first axis₁The second axis₂The interference force τ on the second axis_{twenty one}And compare the behavior (trajectory misalignment) between the real machine and the simulation unit.
Mutual inertia J of the mutual interference compensation unit 35 of the actual machine part_r12(= J_r21) And the mutual inertia J actually acting on the mechanical part of the actual machine part_s12(= J_s21) Does not match, a locus shift occurs between the real machine unit and the simulation unit. Here, the mutual inertia parameter of the mutual interference compensator 35 is adjusted by quantitatively judging the locus deviation between the actual machine part and the simulation part.
[0039]
In order to compare the trajectory between the actual machine unit and the simulation unit, the forward conversion unit 34 includes the position FB of the position detectors 14a and 14b of the motor control unit 11 and the position detector models 24a and 24b of the control model unit 21. Each position FB is forward-converted to determine the position FB (state quantity) of the robot tip of the real machine part and the simulation part in the orthogonal coordinate system. The state comparison unit 31 obtains a state amount comparison value by subtracting the two positions FB, and determines whether the evaluation amount calculation unit 32 is within a predetermined threshold value.
If it is within the threshold value, the control parameter (mutual inertia) adjustment is terminated. As a result, the mutual interference actually acting and the mutual interference compensation unit 35 of the actual unit become approximate values. If it is equal to or greater than the threshold, the control parameter adjustment unit 33 controls the control parameter J of the mutual interference compensation unit 35 as shown below.₁₂Repeat until the value is within the threshold.
[0040]
Mutual inertia parameter J₁₂7 (b), the mutual inertia parameter J is calculated from the deviation (ΔX, ΔY) between the trajectory of the actual machine part and the trajectory of the simulation part.₁₂Judge the magnitude and positive / negative direction to change. For this purpose, the shift amount of the trajectory in the Cartesian coordinate system is converted into a shift amount of the joint coordinate system using an inverse matrix of a minute displacement relational expression between the joint coordinate system called the Jacobian and the work coordinate system. The biaxial Jacobian Jacobian can be expressed as:
[0041]
[Formula 6]

[0042]
Where θ₁₂= Θ₁+ Θ₂
And its inverse matrix Jacobean^-1Can be expressed as:
[0043]
[Expression 7]

[0044]
Therefore, the amount of deviation (ΔX, ΔY) of the locus in the orthogonal coordinate system and the angle error (Δθ in the joint coordinate system).₁, Δθ₂) Can be expressed as follows:
[0045]
[Equation 8]

[0046]
Also, from the rigidity of the joint coordinate system, the relationship between the angle error and the torque can be expressed by the following equation.
[0047]
[Equation 9]

[0048]
Where K = Kp · Kv · J
Therefore, the mutual inertia parameter J₁₂Can be obtained by solving equation (13) and the following equation.
[0049]
[Expression 10]

[0050]
By repeating the above operation using the position command used in the actual operation, it is possible to adjust the control parameter for interference force compensation due to the mutual interference of the multi-axis robot.
[0051]
【The invention's effect】
  As described above, according to the motor control device of the first aspect, the actual machine unit and the simulation unit are arranged in parallel, and the state quantities of the mechanism unit and the mechanism unit model are compared according to the same input command. By repeating the change of the control parameter value of the actual unit from the comparison of the operation by the same command until the state quantity enters within the predetermined threshold, the operation area can be limited without being affected by gravity, dynamic friction, etc. The control parameters of the control unit can be automatically adjusted without being restricted by the above.
  According to the motor control apparatus of the second aspect, the control parameter generated between the axes in the case of a plurality of axes can be adjusted by performing the adjusting operation with the plurality of state amounts of the motor control.Further, by compensating for the mutual interference of the plurality of motors and adjusting the mutual interference parameters, the control parameters of the control unit can be automatically adjusted without being restricted by the mutual interference.
  According to the motor control device of claim 3, the gain adjustment of the motor control unit is performed so that the vibration characteristics of the speed and acceleration of the motor control unit are below a certain level compared to the same signal of the simulation unit. By adjusting, it is possible to accurately determine the model parameters and accurately identify the model parameters regardless of the operating conditions.
  According to the motor control device of claim 4, the vibration characteristics are evaluated accurately by using a technique for evaluating the magnitude of the vibration frequency component such as FFT to accurately determine the vibration component in question. The control parameters can be adjusted.
  According to the motor control device of the fifth aspect, after the evaluation of the vibration characteristics, the control system suitable for the mechanism configuration of the actual machine and the model can be constructed by changing the control structure of the motor control unit. The target can be controlled more precisely.
  Claim6According to the motor control device described,The parameter can be automatically updated by selecting execution of the adjustment operation based on the judgment of the evaluation amount calculation unit.
  Claim7According to the motor control device described above, by setting the control parameter of the motor control unit to the position loop gain or the speed loop gain, highly accurate positioning and speed response are possible.
  Claim8According to the described motor control method, the real machine unit and the simulation unit are arranged in parallel, the state quantities of the mechanism unit and the mechanism unit model are compared according to the same input command, and a predetermined threshold value is determined. Even if there are multiple peripheral devices around the mechanism unit without being affected by gravity, dynamic friction, etc., by repeating the change of the control parameter value of the actual machine unit from the operation by the same command until the state quantity enters within The control parameters of the control unit can be easily adjusted without being restricted by the limitation of the area.
  According to the motor control apparatus of the ninth aspect,By performing the adjustment operation with a plurality of motor control state quantities, the control parameters generated between the axes in the case of a plurality of axes can also be adjusted. Further, by compensating for the mutual interference of the plurality of motors and adjusting the mutual interference parameters, the control parameters of the control unit can be automatically adjusted without being restricted by the mutual interference.
[Brief description of the drawings]
FIG. 1 is a first basic configuration diagram of the present invention.
FIG. 2 is a block diagram showing a first specific embodiment of the present invention.
FIG. 3 shows a response waveform 1 in the first specific embodiment of the present invention.
FIG. 4 is a response waveform 2 in the first specific embodiment of the present invention.
FIG. 5 shows a response waveform in the second specific example of the present invention.
FIG. 6 is a block diagram showing a third specific example of the present invention.
FIG. 7 is a model representing a fourth specific embodiment of the present invention.
FIG. 8 is a block diagram showing a fourth specific example of the present invention.
FIG. 9 is a diagram showing a conventional control method
[Explanation of symbols]
10: Actual machine part
11: Motor controller
12: Mechanism part
13: Motor
14: Position detector
15: External force compensation section
20: Simulation section
21: Control model part
22: Mechanism model part
23: Motor model
24: Position detector model
25: External force compensator model
31: State quantity comparison unit
32: Evaluation amount calculation unit
33: Control parameter adjustment unit
34: Parameter identification unit
35: Mutual interference compensator
111: Position velocity loop
112: Amplifier
121: Secondary mechanism
122: Reducer
211: Position / speed control loop model
212: Amplifier model
221: Secondary mechanism model
222: Reducer model
311: Vibration component extraction unit

Claims (9)

In a motor control device that drives a motor based on a command from a motor control unit and has a mechanism unit connected to the motor,
A simulation unit including a motor model simulating the motor, a mechanism model unit simulating the mechanism unit, and a control model unit simulating the motor control unit;
The motor control unit, the motor and the mechanism unit, and the simulation unit are compared with any one of the motor position, the motor speed, and the motor acceleration, which are the same state quantities , and extracted using FFT A state quantity comparison unit that creates a state quantity comparison value that is a vibration component of the state quantity;
The state quantity comparison value is compared with a preset threshold, and when the state quantity comparison value is equal to or greater than the threshold, it is determined that the control parameter of the motor control unit is not optimal, and the state quantity comparison value An evaluation amount calculation unit that determines that the control parameter is optimal when is within the threshold;
A control parameter adjustment unit that adjusts the control parameter so that the state quantity comparison value falls within the threshold when the evaluation parameter calculation unit determines that the control parameter is not optimal. A motor control device. A motor control device that drives a motor based on a command from a motor control unit and has a mechanism unit connected to the motor, wherein the motor control device drives a multi-axis robot to which a plurality of the motors are connected.
An actual machine unit that includes the motor control unit, the motor, and the mechanism unit, and further includes a mutual interference compensation unit having a mutual inertia parameter for compensating for mutual interference between the shafts;
A simulation unit having a motor model simulating the motor, a mechanism model unit simulating the mechanism unit, and a control model unit simulating the motor control unit for each axis;
A state quantity comparison unit that compares the same state quantities of the real machine unit and the simulation unit and creates a state quantity comparison value;
The state quantity comparison value is compared with a preset threshold value. When the state quantity comparison value is equal to or greater than the threshold, it is determined that the mutual inertia parameter is not optimal. An evaluation amount calculation unit that determines that the mutual inertia parameter is optimal when within the threshold;
A control parameter adjustment unit that adjusts the mutual inertia parameter so that the state quantity comparison value falls within the threshold when the evaluation amount calculation unit determines that the mutual inertia parameter is not optimal. The motor control apparatus characterized by the above-mentioned.   The motor control device according to claim 2, wherein the state quantity is one of a motor position, a motor speed, and a motor acceleration.   The motor control device according to claim 2, wherein the state quantity comparison unit uses a vibration component of the state quantity extracted using FFT as the state quantity comparison value. Even if the control parameter adjustment unit adjusts the control parameter, if the state quantity comparison value is greater than or equal to the threshold,
The motor control device according to claim 1, wherein the motor control device changes a control structure of the motor control unit, a model structure of the mechanism model unit, and a control structure of the simulation unit.   The evaluation amount calculation unit monitors the state quantity comparison value and automatically starts adjusting the control parameter or the mutual inertia parameter according to the magnitude of the state quantity comparison value. Or the motor control apparatus of 2.   The motor control device according to claim 1, wherein the control parameter is a position loop gain or a speed loop gain. A control method for a motor control device that drives a motor based on a command from a motor control unit, includes a mechanism unit connected to the motor, and includes an actual machine unit including the motor control unit, the motor, and the mechanism unit. In
The motor control device further includes a simulation unit having a model simulating the real machine unit,
The same command is input to the actual machine unit and the simulation unit,
A state quantity comparison value that is a vibration component of the state quantity extracted using FFT by comparing any one of the motor position, the motor speed, and the motor acceleration that are the same state quantity in the real machine section and the simulation section. Create
The state quantity comparison value is compared with a preset threshold, and when the state quantity comparison value is equal to or greater than the threshold, it is determined that the control parameter of the motor control unit is not optimal, and the state quantity comparison value Determines that the control parameter is optimal when the threshold is within the threshold;
When the control parameter is determined to be not optimal, the control of the motor control device is processed by a procedure of adjusting the control parameter so that the state quantity comparison value falls within the threshold value. Method. A motor unit is driven based on a command from the motor control unit, and has a mechanism unit connected to the motor, and includes an actual machine unit including the motor control unit, the motor, and the mechanism unit, and a plurality of the motors are connected. In a control method of a motor control device for driving a multi-axis robot,
The motor control device further includes a mutual interference compensation unit having a mutual inertia parameter for compensating mutual interference between axes in the real machine unit, and a simulation unit having a model simulating the real machine unit for each axis. Prepared,
The same command is input to the actual machine unit and the simulation unit,
Compare the same state quantity of the actual machine part and the simulation part to create a state quantity comparison value,
The state quantity comparison value is compared with a preset threshold value. When the state quantity comparison value is equal to or greater than the threshold, it is determined that the mutual inertia parameter is not optimal. Determining that the mutual inertia parameter is optimal when within the threshold;
When it is determined that the mutual inertia parameter is not optimal, the motor control device is processed by a procedure of adjusting the mutual inertia parameter so that the state quantity comparison value falls within the threshold value. Control method.

Images