Servo system fractional order model identification method with time delay link taken into consideration

Abstract

The invention discloses a servo system fractional order model identification method with a time delay link taken into consideration. The method comprises steps as follows: establishing a to-be-identified fractional order model of a servo system with a time delay factor taken into consideration; acquiring frequency domain characteristics of the servo system; constructing a model identification target function; calculating a model parameter matrix of the servo system; determining the optimal fractional order to obtain the fractional order meeting a set threshold value; determining a factor of the time delay link. The fractional order system model with the time delay link is constructed for the servo system, model parameters are acquired by use of frequency domain response data of the system,and the modeling and parameter identification accuracy of the servo system is improved.

Description

technical field [0001] The invention belongs to the technical field of servo control, in particular to a fractional order model identification method of a servo system considering the delay link. Background technique [0002] Servo systems are widely used in advanced intelligent manufacturing fields such as industrial robots and numerical control systems. The motion control performance of the servo system will directly affect the operating efficiency and positioning accuracy of the mechatronics equipment. In order to realize the high-performance motion control of the servo system, precise mathematical modeling is required. [0003] The traditional model identification method first determines the structure of the system model, and then uses the collected data to determine the model parameters through the system excitation of the servo system. The current traditional model identification method has three deficiencies: (1) The mathematical model determined in the traditional ...

AI Technical Summary

Problems solved by technology

The current traditional model identification method has three deficiencies: (1) The mathematical model determined in the traditional method does not consider the fractional order characteristics existing in the system

In the traditional servo system, the integer-order model is used. The premise is that the inertia and friction of the motor are ideal, that is, the load inertia is evenly distributed, and the friction is strictly proportional to the speed. However, In the actual process, the mass of the load is difficult to distribute evenly, and the friction force cannot be strictly proportional to the speed, which will cause the system to produce fractional-order characteristics; (2) Most of the traditional methods use time-domain signals to identify model parameters, and there is no Determine the model structure and obtain the model parameters from the perspective of the frequency domain; (3) When modeling the servo system in the current data, in order to simplify the model, the delay link is mostly ignored, and the actual servo system cannot be modeled. Accurately described, so the control rate designed based on this model has limited dynamic following performance

Images