JPH06119005A - Automatic tuning method for servo system gain using fuzziness - Google Patents

Abstract

PURPOSE:To obtain the automatic gain tuning of a servo system which can execute tuning again even if a motor is in the middle of a stop from information on the vibration of a machine system even when tuning is executed at an unstable area through the use of a standard servo motor. CONSTITUTION:In the servo system composed of the positioning unit and the like A of PC outputting the command of operation speed and the like, a fuzzy inference part B, a servo amplifier C and the inexpensive and standard servo motor D, the amplitude of current is fetched from a current sensor C5 included in the servo amplifier C by a motor current amplitude detector B1 as the input element of a fuzzy inference module B4 and it is inferred. Thus, the automatic gain tuning of the servo system is executed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of automatically tuning a gain of a servo system by using fuzzy to perform gain tuning of the servo system by fuzzy.

[0002]

2. Description of the Related Art FIG. 2 shows an example of a conventional fuzzy auto-tuning configuration. In the figure, A is a positioning unit of the PC for issuing commands such as operating speed to the servo amplifier. B is a fuzzy inference unit configured by PC peripherals and the like for performing fuzzy tuning calculation.
Is a overshoot detector that detects the amount of overshoot from the speed information, B3 is a settling time detector that detects the settling time, and B4 is a fuzzy inference module that infers the optimum gain from the information from these B2 and B3. C is a servo amplifier, C1 is a position loop gain representing a position feedback gain, C2 is a speed control gain representing a speed feedback gain, C3 is a current control gain representing a current feedback gain, C4 is a motor driving inverter, and C5 is Current detector for current feedback, C
Reference numeral 6 is an interface for connecting to the pulse generator of the servo motor. Further, D represents an inexpensive standard servomotor instead of a special servomotor with a vibration sensor, in which D1 is a motor section that produces power, and D2 is a pulse generator section that outputs the motor rotation speed.

When performing auto-tuning of a servo system using fuzzy, the fuzzy inference module infers the magnitude of each gain from the overshoot of the input element obtained from the speed information and the settling time, and for example, the overshoot. If is small or the settling time is long, the gain is increased, and if the overshoot is large, the gain is decreased.
Tuning is performed in this manner, and good tuning is normally performed. However, in reality, the servo motor is often overloaded, and in this case, even if the tuning result is in the unstable area,
In a range where the vibration is small, mechanical vibrations are less likely to appear in overshoot, settling time, etc., so unstable information is difficult to convey to the inference module side. In addition, even when considering installing a vibration sensor (acceleration sensor) in the mechanical system to prevent this, the effect cannot be expected until it enters an unstable state where vibration appears in the mechanical system when overloaded, and It leads to cost increase.

[0004]

As described above, in the gain tuning of the servo system using the fuzzy control based on the prior art, even when it is difficult to appear in the overshoot or the settling time even in the unstable region. It was difficult to convey the unstable state to the inference module side, and there were times when tuning was performed as a stable state.

The present invention has been made in order to solve the above problems, and even when tuning to an unstable region for some reason, the tuning is performed even when the motor is stopped by judging from the vibration information of the mechanical system. The purpose is to start over. In addition, the current information that flows in the motor is used as a substitute for the mechanical system vibration information, and the current sensor output included in the servo amplifier is used. The purpose is to enable the use of servo motors.

[0006]

A method of auto-tuning a gain of a servo system using fuzzy according to the present invention uses a fuzzy control for gain adjustment of an amplifier in a servo control system including a servo amplifier and a servo motor. The motor-side vibration element is used as one of the input elements of the inference module.

[0007]

The information from the current sensor used for current feedback in the servo amplifier is used as the vibration element of the mechanical system. The amplitude of the current is obtained from this and used as one of the input signals to the fuzzy inference module. The fuzzy inference module infers from these input elements as described above and suppresses the gain when the vibration of the mechanical system is large. Further, since the vibration factor is obtained from the electric current, it is not necessary to attach a sensor to the motor side.

[0008]

【Example】

Example 1. FIG. 1 is a block diagram of an embodiment of a fuzzy auto-tuning method according to the present invention. In the figure, A is a positioning unit of the PC for issuing commands such as operating speed to the servo amplifier. Reference numeral B is a fuzzy inference unit for calculating tuning, which is composed of PC peripherals, B1 is a motor current amplitude detector for obtaining an amplitude from current information, and B2 is an overshoot for detecting an overshoot amount from speed information. A detector, B3 is a settling time detector that detects the settling time, and B4 is an inference module that infers an optimum gain from the information from these B2 and B3. C is a servo amplifier, C1 is a position loop gain showing a position feedback gain, C2 is a speed control gain showing a speed feedback gain, C3 is a current control gain showing a current feedback gain, C4 is a motor drive inverter, and C5 is Current detector for current feedback,
C6 is an interface for connecting to a pulse generator of a servo motor described later. Further, D is an inexpensive general-purpose servomotor, D1 is a motor unit that produces power, and D2 is a pulse generator unit that outputs the motor rotation speed.

The operation of the present invention will be described with reference to FIG. It is assumed that the tuning state has entered the unstable region for some reason. In this case, the vibration of the mechanical system is large, and the value of the current from the current sensor included in the servo amplifier is greatly vibrating. The amplitude of the current is taken out by the motor current amplitude detector and input to the input of the fuzzy inference module as vibration information of the mechanical system. Based on the rules, the fuzzy inference module works to increase the gain when the vibration of the mechanical system is small, and works to decrease the gain when the vibration of the mechanical system is large. Due to this function, good tuning with a sufficiently large gain can be obtained within an allowable range even for an increase in vibration that does not appear in data such as overshoot and settling time. Further, since the vibration appears more sensitively to the current information than the direct vibration information, especially when overloaded, by using the current vibration of the motor for the vibration information of the mechanical system, the allowable range can be obtained even in the small vibration area. Can be set.

By the way, in the above description, the present invention is applied to the automatic tuning of the gain of the servo motor by the fuzzy method. For example, the amplitude information of the current is used as the mechanical vibration information only to reduce the acceleration sensor. It can also be applied to such uses.

[0011]

As described above, according to the present invention, it is possible to perform gain tuning in consideration of the vibration width of the mechanical system.

Further, by using the motor current information from the current sensor included in the servo amplifier and used for normal current feedback, there is an effect that it can be realized by using an inexpensive standard servomotor. .

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration of a fuzzy auto gain tuning of a servo system according to a first embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of gain auto-tuning of a servo system by fuzzy in the prior art.

[Explanation of symbols]

A Positioning unit of PC (programmable controller) B Fuzzy inference section B1 Motor current amplitude detector B2 Overshoot detector B3 Settling time detector B4 Fuzzy inference module C Servo amplifier C1 Position feedback loop gain C2 Velocity feedback loop gain C3 Current feedback Loop gain C4 Inverter C5 Current sensor C6 Interface for pulse generator D Servo motor D1 Motor part D2 Pulse generator part

Claims (1)

[Claims] 1. In a servo control system including a servo amplifier and a servo motor, fuzzy control is used for gain adjustment of the amplifier, and a vibration element on the motor side is used as one of the input elements of the inference module. Characteristic servo auto gain tuning method.