Position control device

Abstract

To provide a position control device that prevents stick-slip.SOLUTION: A position control device includes an integral time constant calculation unit 5 that notifies a speed control unit 2 of an integral time constant. If a stick-slip occurs in which the value of the integral time constant changes continuously while the function generation is stopped, the integral time constant calculation unit 5 does not switch the value of the integral time constant even when a position deviation absolute value |ΔP| crosses a switching threshold INP in the decreasing direction, but switches the value of the integral time constant when the position deviation absolute value |ΔP| crosses an additional threshold INP1 smaller than the switching threshold INP in the decreasing direction and when it crosses the switching threshold INP in the increasing direction.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a position control device that controls a feed axis of a machine tool or the like using a servo motor. [Background technology]

[0002] The configuration of a commonly used position control device and the flow of various signals are explained with reference to FIG.

[0003] A position command value P* is input from a higher-level control device (not shown). Subtractor 56 subtracts position feedback P from position command value P* to calculate position deviation ΔP (=P*-P). Position feedback P indicates a position detection signal that detects the position of the control target of target system 54. Position deviation ΔP is proportionally amplified by a position loop gain or the like in position control unit 51 to become speed command value V*. Subtractor 57 subtracts speed feedback V from speed command value V* to calculate speed deviation ΔV (=V*-V). Speed ​​feedback V is the time differential value of the rotational angle position of a position detector (not shown) connected to the servo motor, or the output of a speed detector (not shown) connected to the servo motor.

[0004] The integral time constant calculation unit 55 calculates an integral time constant based on the function generation stop signal and the position deviation ΔP, and notifies the speed control unit 52 of the calculated integral time constant.

[0005] The speed control unit 52 performs PI control or the like on the speed deviation ΔV using the integral time constant notified by the integral time constant calculation unit 55, and calculates a torque command value T. The torque command value T becomes a generated torque for the target system 54 in the conversion unit 53.

[0006] In positioning control of feed axes of machine tools using servo motors, a known method for preventing overshoot during positioning is to reduce the value of the velocity loop integrator and lower the servo motor output torque when the position error approaches "0." A widely used method for lowering this output torque is to switch the velocity loop integrator between perfect integration and imperfect integration, i.e., to switch the integral time constant.

[0007] The integral time constant is normally the first integral time constant. However, it switches to the second integral time constant only when the movement command from the upper control device is "0" and the position deviation is below a predetermined value. In the following, when the movement command from the upper control device becomes "0", it is referred to as "function generation stops."

[0008] Furthermore, Patent Document 1 proposes that the integration process of the speed loop be incompletely integrated immediately before the servo motor stops rotating, and that the integration process of the speed loop be completely integrated while the servo motor is normally running. The technology of Patent Document 1 maintains servo stiffness and reduces the occurrence of overshoot when the servo motor stops. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] Patent No. 3197898 Summary of the Invention [Problem to be solved by the invention]

[0010] When control that switches between perfect integration and imperfect integration is adopted in the integral process, the value of the integral time constant is switched when the position deviation falls below a predetermined value after the function generation from the upper control device stops. An example of switching the integral time constant is shown in Fig. 7. In Fig. 7, when the absolute value |ΔP| of the position deviation ΔP falls below a specified switching threshold INP, the value of the integral time constant becomes the second time constant G2, and when the absolute value |ΔP| of the position deviation exceeds the switching threshold INP, the value of the integral time constant becomes the first time constant G1.

[0011] Here, referring to FIG. 6, we will explain the phenomenon that occurs when the integral time constant is switched based on the absolute value of the position deviation |ΔP|. During positioning to the target value, when the absolute value of the position deviation |ΔP| becomes equal to or less than the switching threshold value INP at time t1, the time constant is switched to the second time constant G2. This gradually decreases the integral value. When the torque command value subsequently becomes less than the external force, including the sliding torque, the position cannot be maintained. As a result, at time t2, the absolute value of the position deviation |ΔP| becomes greater than the switching threshold value INP, and the integral time constant is switched to the first time constant G1. With the switch to the first time constant G1, the integral value and the torque command value gradually increase. When the torque command value exceeds the external force, including the sliding torque, movement toward the target value begins. Then, at time t3, when the absolute value of the position deviation |ΔP| becomes equal to or less than the switching threshold value INP, the time constant is switched back to the second time constant G2. Subsequently, the same process is repeated, resulting in stick-slip, in which the object alternates between moving and stopping. It should be noted that G1 and G2 in FIG. 6 indicate the values ​​of the time constants used when the integral time constants are switched as shown in FIG.

[0012] As described above, in the conventional technology, the integral time constant is repeatedly switched, which results in the occurrence of stick-slip. [Means for solving the problem]

[0013] A position control device disclosed in this specification is a position control device that issues commands to a motor that drives a controlled system and controls the position of the controlled object, and includes: a position control unit that performs position control based on a position command value and position feedback of the controlled object; a speed control unit that performs speed control based on a speed command value and speed feedback of the controlled object; and an integral time constant calculation unit that notifies the speed control unit of an integral time constant, wherein the integral time constant calculation unit switches the value of the integral time constant when, in a function generation stopped state in which an instruction to stop movement of the controlled object has been received, the absolute value of a position deviation, which is the difference between the position command value and the position feedback, crosses a predetermined switching threshold, and is characterized in that, when stick-slip occurs in the function generation stopped state in which the value of the integral time constant is continuously switched, the integral time constant calculation unit does not switch the value of the integral time constant even if the absolute value of the position deviation crosses the switching threshold in a decreasing direction, but switches the value of the integral time constant when the absolute value of the position deviation crosses an additional threshold that is smaller than the switching threshold in a decreasing direction and when it crosses the switching threshold in an increasing direction. [Effects of the Invention]

[0014] By adding a new threshold for switching the integral time constant and providing hysteresis when switching the integral time constant, it is possible to suppress overshoot during positioning, prevent stick-slip, and perform stable control. [Brief explanation of the drawings]

[0015] [Figure 1] 1 is a diagram showing the configuration of a position control device according to an embodiment of the present invention and the flow of various signals; [Figure 2] FIG. 2 is a diagram illustrating details of the configuration of a speed control unit in the position control device according to one embodiment of the present invention. [Figure 3] 10A and 10B are diagrams illustrating the behavior of the position of a control object during positioning in the present invention. [Figure 4] FIG. 10 is a diagram illustrating a method for switching an integral time constant according to the present invention. [Figure 5]FIG. 1 is a diagram illustrating the configuration of a conventional position control device that uses a system for switching integral time constants and the flow of various signals. [Figure 6] FIG. 10 is a diagram illustrating the behavior of the position of a control object during positioning using a conventional method. [Figure 7] FIG. 10 is a diagram illustrating a conventional method for switching integral time constants. DETAILED DESCRIPTION OF THE INVENTION

[0016] FIG. 1 is a diagram showing an example of the configuration of a position control device according to an embodiment of the present invention and the flow of various signals, and FIG. 2 is a diagram showing the details of the configuration of the speed control unit 2. As shown in FIG.

[0017] The configuration of a position control device according to an embodiment of the present invention and the flow of various signals will be described with reference to Figure 1. A position command value P* is input to the position control device from a higher-level control device (not shown). A subtractor 11 subtracts a position feedback P from the position command value P* to calculate a position deviation ΔP (=P*-P). The position feedback P indicates a position detection signal that detects the position of the control target of the target system 4. The position deviation ΔP is proportionally amplified by a position loop gain or the like in the position control unit 1 to become a speed command value V*. A subtractor 12 subtracts the speed feedback V from the speed command value V* to calculate a speed deviation ΔV (=V*-V). The speed feedback V is the time differential value of the rotational angle position of a position detector (not shown) connected to the servo motor, or the output of a speed detector (not shown) connected to the servo motor.

[0018] The integral time constant calculation unit 5 includes a determination unit 6, a correction unit 7, and a calculation unit 8. The integral time constant calculation unit 5 notifies the speed control unit 2 of the integral time constant based on a function generation stop signal notified from a higher-level control device (not shown), the position deviation ΔP, and the position feedback P. Here, the function generation stop signal has been described as being notified from a higher-level control device (not shown), but it is also possible to determine whether the signal is on or off from a movement command from the higher-level control device or a position command value P*.

[0019] The determination unit 6 determines whether or not to add a threshold for switching the integral time constant based on the function generation stop signal and the position feedback P, and notifies the calculation unit 8 of the result. The position feedback P is a position detection signal that detects the position of the control target of the target system 4. Furthermore, the correction unit 7 calculates a corrected position deviation ΔPC for the position deviation ΔP based on the switching threshold from the calculation unit 8, and notifies the calculation unit 8 of the result together with the position deviation ΔP. Here, if no correction is performed, ΔPC = ΔP. Furthermore, the calculation unit 8 calculates the integral time constant based on the threshold notified from the determination unit 6 and the position deviation ΔPC notified from the correction unit 7.

[0020] The speed control unit 2 performs PI control or the like on the speed deviation ΔV using the integral time constant notified by the integral time constant calculation unit 5, and calculates a torque command value T. The torque command value T becomes the torque to be generated for the target system 4 in the conversion unit 3.

[0021] Next, the speed control unit 2 will be described with reference to Fig. 2. Amplifier 21 proportionally amplifies the speed deviation ΔV with a speed loop gain Q. Furthermore, integral calculation unit 27 calculates an integral component from the speed deviation ΔV. Here, in integral calculation unit 27, amplifier 22 proportionally amplifies the speed deviation ΔV with a speed loop gain I. Adder 25 adds together a previous value 24 of the integral component and the output value of amplifier 22. The output value of adder 25 is multiplied by integral time constant G. Adder 26 outputs the sum of the output value of adder 25, i.e., the output value of integral calculation unit 27 and the proportionally amplified proportional component, as a torque command value T. Furthermore, integral time constant G is an integral time constant notified from integral time constant calculation unit 5.

[0022] Next, the determination unit 6, correction unit 7, and calculation unit 8 that make up the integral time constant calculation unit 5 will be described. The determination unit 6 determines whether or not stick-slip has occurred. Specifically, the determination unit 6 confirms that function generation has stopped from the function generation stop signal, and determines that stick-slip has occurred if the position deviation absolute value |ΔP| exceeds the switching threshold value INP again after falling below the switching threshold value INP. If it is determined that stick-slip has occurred, the determination unit 6 further calculates an additional threshold value INP1 using the following formula from the difference ΔX (=P1-P2) between position feedback P1 and position feedback P2, which will be described later, and notifies the calculation unit 8 of the result. INP1 = INP - |ΔX| (Equation 1)

[0023] FIG. 3 is a diagram showing an example of the behavior of the position of the controlled object during positioning when stick-slip occurs and an additional threshold INP1 is added.

[0024] Position feedback P1 indicates the position feedback when the controlled object is stationary with the position deviation absolute value |ΔP| equal to or less than the switching threshold INP, and position feedback P2 indicates the position feedback when the controlled object is stationary with the position deviation absolute value |ΔP| exceeding the switching threshold INP. Also, G1 and G2 in Fig. 3 indicate the values ​​of the time constants used when the integral time constants are switched as shown in Fig. 4.

[0025] Here, when |ΔX| is large and the additional threshold INP1 has a negative value, the relationship between the additional threshold INP1 and the position command value P* is reversed, and therefore the position command value P* needs to be corrected.

[0026] In the correction unit 7, when the additional threshold INP1 is a positive value, no correction is made, so that ΔPC = ΔP. When the additional threshold INP1 is a negative value, the position command value P* is corrected by K × (|ΔX|-INP) using a proportionality constant K (K > 1) so that the relationship between the additional threshold INP1 and the position command value P* is not reversed. Therefore, the corrected position deviation ΔPC is calculated from the position deviation ΔP based on the following formula, and is notified to the calculation unit 8 together with the position deviation ΔP. ΔPC=P*+SGN(ΔX)×K×(|ΔX|-INP)-P =ΔP+SGN(ΔX)×K×(|ΔX|-INP) (Formula 2) In Equation 2, SGN(ΔX) is a sign function and has the following value: SGN(ΔX) = 1(ΔX>0) 0 (ΔX=0) -1(ΔX<0)

[0027] When correcting the position command value P*, the switching threshold value INP and the additional threshold value INP1 are corrected by the amount of correction of the position command value P* using the following formula. INPC = INP + K × (|ΔX|- INP) (Equation 3) INP1C=INP1+K×(|ΔX|-INP) (Formula 4)

[0028] Furthermore, INP1C is expressed by the following equation using (Equation 1), and since K>1 and (|ΔX|−INP)>0, it is corrected to a positive value. INP1C=INP-|ΔX|+K×(|ΔX|-INP) =(K-1)(|ΔX|-INP) (Equation 5)

[0029] Here, in the case of positioning as shown in Fig. 3, when correction is made to the position command value P*, |ΔP|, INP, and INP1 in Fig. 4 are replaced with |ΔPC|, INPC, and INP1C, respectively, and the integral time constant is switched. In other words, when INP1<0, the corrector 7 corrects |ΔPC|, INPC, and INP1C so that they increase by K×(|ΔX|-INP) compared to |ΔP|, INP, and INP1.

[0030] Furthermore, the position deviation ΔP input to the position control unit 1 uses the position deviation ΔPC notified by the corrector 7 so that the relationship between INP1 and the position command value P* is not reversed.

[0031] The calculation unit 8 calculates the integral time constant G based on the threshold value INP1 notified by the determination unit 6 and the position deviation ΔPC notified by the correction unit 7, based on FIG.

[0032] FIG. 4 is a diagram showing a method of switching the integral time constant when the additional threshold value INP1 is added, and shows an example of the relationship between the position deviation absolute value |ΔP| and the integral time constant G.

[0033] As shown in FIG. 4, the integral time constant G normally uses the first integral time constant G1, but when the position error absolute value |ΔP| gradually decreases while the function generation is stopped. Then, when |ΔP| becomes equal to or less than the additional threshold value INP1, the integral time constant G is switched to the second integral time constant G2. Furthermore, after switching to the second integral time constant G2, if the position error absolute value |ΔP| increases due to an external force or the like and exceeds the switching threshold value INP, the integral time constant G is switched from the second integral time constant G2 to the first integral time constant G1. On the other hand, even if the position error absolute value |ΔP| crosses the switching threshold value INP during a decrease, or crosses the additional threshold value INP1 during an increase, the integral time constant G is not switched. In other words, when the absolute value of the position error |ΔP| is within the range from INP1 to INP, the integral time constant G will have multiple values ​​(G1 or G2) depending on the direction of increase or decrease of the position error |ΔP|. In other words, the integral time constant G is switched by providing hysteresis.

[0034] FIG. 3 is a diagram illustrating an example of the behavior of the position of the controlled object when an additional threshold value INP1 is set. As shown in FIG. 3, after the function generation is stopped, at time t1, |ΔP|≦INP holds true, and therefore the time constant is switched to the second integral time constant G2 at time t1. Thereafter, at time t2, |ΔP|>INP holds true, and therefore the time constant is switched to the first integral time constant G1 at time t2. After |ΔP|≦INP holds true at time t1, |ΔP|>INP holds true at time t2, and therefore the determination unit 6 determines that stick-slip has occurred. In this case, the determination unit 6 sets an additional threshold value INP1 in addition to the normal switching threshold value INP as a threshold value for switching the time constant. Therefore, after time t2, the time constant is not switched even if |ΔP|≦INP holds true, and the time constant is switched only after |ΔP|≦INP holds true. 3, at time t3, the time constant switches from the first integral time constant G1 to the second integral time constant G2. With this configuration, as shown in FIG. 3, the position deviation absolute value |ΔP| can be prevented from exceeding the switching threshold value INP, thereby preventing stick-slip.

[0035] Here, when the function generation stop signal is turned off, i.e., when function generation starts, the additional threshold INP1 is deleted, and at the same time, the position deviation ΔPC becomes the pre-correction position deviation ΔP, and the integral time constant is switched according to FIG. 7.

[0036] Furthermore, if the position feedback P is not stable even after adding the additional threshold value INP1, the position deviation absolute value |ΔP| does not become equal to or less than the switching threshold value INP, and stick-slip occurs, it is possible to update the additional threshold value INP1 again. In this case, the updated additional threshold value INP1 will be even smaller than the pre-update additional threshold value INP1.

[0037] In this way, by adding the additional threshold value INP1 and switching the integration time constant G, the position feedback P exhibits the behavior as shown in FIG. 3. That is, positioning can be achieved without causing stick-slip and with |ΔP| < INP. Also, since the torque decreases before the final positioning, the torsional component of the target system is eliminated, resulting in a stable state in terms of control.

[0038] In the above embodiment, when |ΔX| is large and INP1 becomes a negative value, the relationship between INP1 and the position command value P* is reversed, so it was described that the position deviation ΔP is changed to an internally corrected value. However, if INP1 is not a negative value and the position feedback P does not fall within INP even when the integration time constant is switched as shown in FIG. 4, it is also possible to change the position deviation ΔP to an internally corrected value.

Explanation of Signs

[0039] 1,51 Position control unit 2,52 Speed control unit 3,53 Conversion unit 4,54 Target system 5,55 Integration time constant calculation unit 6 Judgment unit 7 Correction unit 8 Calculation unit 11,12,56,57 Subtractor 21,22 Amplifier 23 Integration time constant 24 Previous value of integral component 25,26 Adder 27 Integral operation unit

Claims

1. A position control device that issues a command to a motor that drives a controlled system and controls the position of the controlled system, a position control unit that performs position control based on a position command value and position feedback of the controlled object; a speed control unit that performs speed control based on a speed command value and a speed feedback of the controlled object; an integral time constant calculation unit that notifies the speed control unit of an integral time constant, and that switches the value of the integral time constant when an absolute value of a position deviation, which is a difference between the position command value and the position feedback, crosses a predetermined switching threshold value in a function generation stop state, which is a state in which an instruction to stop movement of the controlled object has been received; Equipped with the integral time constant calculation unit, when a stick-slip occurs in which the value of the integral time constant is continuously switched, does not switch the value of the integral time constant even if the absolute value of the position deviation crosses the switching threshold in a decreasing direction while in the function generation stopped state, and switches the value of the integral time constant when the absolute value of the position deviation crosses an additional threshold smaller than the switching threshold in a decreasing direction and when the absolute value of the position deviation crosses the switching threshold in an increasing direction. A position control device characterized by:

2. 2. The position control device according to claim 1, wherein the integral time constant calculation unit determines that the stick-slip has occurred when, after the function generation stopped state, the absolute value of the position deviation becomes equal to or less than the switching threshold and then exceeds the switching threshold again.

3. 2. The position control device according to claim 1, a position control device, characterized in that the integral time constant calculation unit calculates the additional threshold value based on a position feedback when the controlled object is stationary with the absolute value of the position deviation being equal to or less than the switching threshold value, a position feedback when the controlled object is stationary with the absolute value of the position deviation exceeding the switching threshold value, and the switching threshold value.

4. 3. The position control device according to claim 2, The integral time constant calculation unit is characterized in that, when the additional threshold is INP1, the position feedback when the controlled object is stationary with the absolute value of the position deviation being equal to or less than the switching threshold is P1, the position feedback when the controlled object is stationary with the absolute value of the position deviation exceeding the switching threshold is P2, and the switching threshold is INP, INP1=INP-|P1-P2| is satisfied.

5. 5. A position control device according to claim 1, a position control device characterized in that, when the absolute value of the position deviation exceeds the switching threshold again after becoming equal to or less than the additional threshold, the integral time constant calculation unit switches the value of the additional threshold to an even smaller value.

Images