A parameter identification method of ball club model based on weighted multi-innovation stochastic gradient

Abstract

The invention discloses a parameter identification method of a club model based on weighted multi-innovation random gradient, which comprises the following steps: collecting single input and output data of the club model according to the inclination angle of the club and the position of the ball; Based on the single input and output data, the ball club model is estimated to be the output error model. The output of the auxiliary model is used to replace the noise-free output of the output error model, and the information vector, the stacking information matrix and the stacking output vector areconstructed. The innovation vector of the club model is calculated based on the information vector, the stacking information matrix and the stacking output vector. Based on the idea of particle weight calculation in particle filter algorithm, the importance of each innovation in the innovation vector is calculated. Calculating a weight of each corresponding innovation according to the importance,and constructing a weight vector; Reasonably assigning the weight of each innovation in the multi-innovation stochastic gradient to identify the parameters of the club model; The invention can realize fast on-line identification, estimate the parameters in the club system model, and has simple calculation method and high efficiency.

Description

technical field [0001] The invention belongs to the field of system identification, and relates to a club system model parameter identification method, in particular to a club model parameter identification method based on weighted multi-innovation stochastic gradient. Background technique [0002] The cue system is a teaching experimental equipment designed for basic control courses in automatic control, mechanical electronics, electrical engineering and other majors. Because the cue system has the characteristics of open-loop instability, it is necessary to design a controller to control the position of the ball, and Its structure is simple, intuitive and clear. It can meet the experimental requirements of courses such as automatic control theory and modern control engineering, and can also be used as experimental equipment for courses such as electrical machinery, motor and drive, and pattern recognition. [0003] The club system is divided into two major components, the...

AI Technical Summary

Problems solved by technology

[0005] The object of the present invention is to provide a cue model parameter identification method based on weighted multi-innovation stochastic gradient for the problems of large amount of calculation, slow calculation speed, weak anti-interference ability to noise, etc. in the model parameter identification of the existing cue system , this method can quickly identify and estimate the parameters in the cue system model online in real time. The specific technical scheme is as follows:

Images