JPH04257001A - Gain scheduling - Google Patents

Abstract

PURPOSE:To smoothly operate a mechanism part whose static friction is large, which is driven by a survo motor. CONSTITUTION:An estimated speed XX1 is searched by an observer 40. The resolution of the estimated speed XX1 is higher than that of a real speed X1. An integral gain K2 of an integral element 33 is changed by this estimated speed XX1. Thus, even the mechanism part whose static friction is large can be smoothly driven.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to gain scheduling in a control system for a mechanism driven by a servo motor, and more particularly to gain scheduling that changes integral gain depending on speed.

0002

2. Description of the Related Art As a method of compensating for nonlinear terms such as static friction in a robot arm driven by a servo motor, there is gain scheduling that changes the gain according to the speed. In robots, the feedback speed is obtained by differentiating the feedback pulses from a pulse coder connected to the servo motor shaft. The gain is changed according to this speed.

0003

However, in such a servo system, if the resolution of the pulse coder is low, the gain cannot be changed precisely and smooth movement of the arm cannot be obtained. The present invention has been made in view of these points, and it is an object of the present invention to provide gain scheduling that changes the gain based on a high-resolution estimated speed estimated by an observer.

0004

[Means for Solving the Problems] In order to solve the above problems, the present invention provides P of a system driven by a servo motor.
Gain scheduling for compensating for static friction caused by an I control loop is characterized in that an estimated speed with high resolution is estimated by an observer, and the gain is changed according to the estimated speed.

[0005]

[Operation] For example, an observer can estimate an estimated speed with a higher resolution than a pulse coder with a lower resolution. Therefore, by changing the gain at this estimated speed, fine-grained gain changes can be made.

[0006]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 2 is a hardware configuration diagram of a robot system for implementing the present invention. host processor 9
is the processor that controls the entire robot. The robot position command is sent from the host processor 9 to the shared RAM 10.
will be written to. Note that the ROM, RAM, etc. coupled to the host processor 9 are omitted. A DSP (digital processor) that controls the servo motor 22 built into the robot.
The signal processor (signal processor) 11 reads the position command in the shared RAM 10 at regular intervals according to the system program in the ROM 12.

The DSP 11 determines a positional deviation based on the difference between this position command and position feedback from a pulse coder 23 built into the servo motor 22. A speed command proportional to this positional deviation is output. Furthermore, the speed of the servo motor 22 is determined by differentiating the position feedback. A speed loop is formed from this speed command and speed. Also,
A current loop is formed by converting the command from the speed loop into a current command. A current command is output from the current loop, from which a PWM waveform for driving the servo motor 22 is generated and sent to the servo amplifier 21 via the DSL (digital servo LSI) 14. Servo amplifier 21 is PWM
Upon receiving the command, the servo motor 22 is driven. Servo motor 22 drives arm 26 via a reduction gear. The servo motor 22 has a built-in pulse coder 23, and feeds back pulses back to the DSP 11 via the DSL 14. Furthermore, the DSP 11 executes an observer function as described later, and estimates the estimated speed, disturbance torque, etc. of the servo motor 22.

FIG. 1 is a block diagram of a servo system for implementing the present invention. The speed command Vc is obtained by subtracting the speed (X1) of the servo motor 22 by an adder 31 and sent to a proportional element 32 and an integral element 33. The proportional element 32 multiplies the gain K1, and the integral element multiplies the integral gain K2 for integration.The respective outputs are added by the adder 34 and output as the current command U1. Current command U1 is input to element 35 having torque constant Kt. The output of element 35 becomes a driving torque that drives servo motor 22. In addition to this, disturbance torque
2 is added by an adder 36. The output of the adder 36 is input to an element 37 having an inertia J, including the servo motor 22, the robot arm 26, and the like. The output of element 37 is acceleration X1(1). This acceleration X1 (1)
is integrated by the integral element 38 and becomes the velocity X1.

On the other hand, the current command U1 and the speed X1 of the servo motor 22 are input to the observer 40. observer 4
0 calculates the estimated speed XX1 from the current U1 and the speed X1 of the servo motor 22. This estimated speed XX1 has higher resolution than the speed obtained from the pulse coder 23. The integral gain K2 is changed based on this estimated speed XX1.

The observer 40 obtains a state equation from the element 35, adder 36, element 37, and integral element 38 in FIG.
From this state equation, estimated speed XX1 and disturbance torque
2, and the estimated speed XX
1 and disturbance torque X2 are estimated. FIG. 3 is a diagram showing this state equation. The observer 40 is configured by state equation (1). Of course, the function of observer 40 is DS
It is executed by P11.

Next, a description will be given of how the integral gain K2 is changed depending on the estimated speed XX1. FIG. 4 is a diagram showing the relationship between estimated speed and integral gain. In the figure, the horizontal axis is the estimated speed XX1, and the vertical axis is the integral gain K2.
It is. As is clear from the figure, when the estimated speed XX1 is 0, the integral gain K2 is maximum, and decreases linearly as the estimated speed changes positive or negative.

[0012] In the above explanation, the mechanical part was explained using a robot arm, but the present invention can be similarly applied to other mechanical parts with large static friction.

[0013]

[Effects of the Invention] As explained above, in the present invention, the gain of the servo system is changed depending on the estimated speed, so the gain can be changed finely based on the estimated speed with high resolution, and the mechanism part can be moved smoothly. Can be done.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of a servo system for implementing the present invention.

FIG. 2 is a hardware configuration diagram of a robot system for implementing the present invention.

FIG. 3 is a diagram showing an equation of state.

FIG. 4 is a diagram showing the relationship between estimated speed and integral gain.

[Explanation of symbols]

9 Host processor 10 Shared RAM 11 DSP (Digital Signal Processor)
14 DSL (Digital Servo LSI) 21
Servo amplifier 22 Servo motor 23 Pulse coder 26 Arm 40 Observer

Claims (3)

[Claims] 1. In gain scheduling for compensating for static friction caused by a PI control loop of a system driven by a servo motor, an estimated speed with high resolution is estimated by an observer, and the gain is changed according to the estimated speed. Gain scheduling characterized by. 2. Gain scheduling according to claim 1, wherein the gain is an integral gain of a speed control loop. 3. The integral gain is controlled to be maximum when the estimated speed is 0, and to decrease linearly in response to positive or negative changes in the estimated speed. gain scheduling.