PID control parameter off-line tuning method based on hidden Markov model

Abstract

The invention relates to a PID control parameter off-line tuning method based on a hidden Markov model. The method comprises the steps of acquiring a PID control parameter observation sequence, judging whether the length of the acquired observation sequence is greater than 1, calculating the initial state probability distribution of the observation sequence, carrying out hidden Markov model training with use of the obtained probability distribution to obtain the state probability of the observation sequence after training, sending the trained control parameter observation sequence to a PID control system for verification, and judging whether the set of control parameters meets the precision requirement of the PID control system. By adopting the method of the invention, deterministic calculation of a PID control parameter sequence vector is realized, an offline PID control parameter optimization adjustment method is proposed, and the difficulty in parameter debugging caused by the randomness of parameter determination is avoided.

Description

technical field [0001] The invention belongs to the field of controller design and relates to an off-line tuning method for PID control parameters based on a hidden Markov model. Background technique [0002] At present, most of the control algorithms used in engineering applications are PID control algorithms. This algorithm has the advantages of simple structure, clear physical meaning, small number of control parameters, and easy engineering debugging. It plays an irreplaceable role in scientific research and production practice. For the determination method of PID control parameters, offline adjustment and online adjustment are mainly used. Off-line adjustment is to lock the range of PID control parameters after the controller form is determined. The off-line optimization process designed in the article "Optimum Fuzzy-GA PID Controller and Its Application" (Tan Guanzheng, Li Anping, Wang Yuechao. Journal of Central South University of Technology, 2002,33(4):419-423.) is...

AI Technical Summary

Problems solved by technology

In the tuning process, the performance index is optimized according to certain optimization rules, and then the PID parameters that meet the requirements are obtained. However, in the tuning process of the PID controller, multiple requirements of the controlled system often need to be considered. The requirements that can be included are limited, and generally cannot completely cover all the requirements. In addition, the existing technology does not use a rigorous mathematical derivation process to describe the tuning process of the PID control parameters, and the accuracy and comprehensiveness are greatly reduced.

Images