Study On Self-Learning Fuzzy Control Strategy For Electro-Hydraulic Servo System

With the rapid development of science research and modern industry, and increasing demands on the servo system's control performance, meanwhile because of many practical systems with characteristics of nonlinear,time-varying parameters,external load disturbance and other factors, it is usually difficult to obtain good control performance using the traditional control method. The intelligent control strategy's research and application may provide some new method to improve the control performance of such complex systems.For performance of the electro-hydraulic servo fuzzy control system needing to be improved, a study on Self-Learning Fuzzy Control strategy is proposed. IN order to improve the dynamic state quality, it is necessary to revise fuzzy rules constantly, adjust the scale factor and perfect fuzzy relations with self-learning fuzzy control algorithm.Firstly, the fuzzy mathematic basis and the basic principles are described. The correct selection of fuzzy control rules,quantitative factors,scale factor and membership functions determines the control performance of the fuzzy control system. However, the understanding of things and the summary of control experiences are ofen subjective, so the establishment of fuzzy control rules and control parameters also has the subjectivity inevitably. In order to eliminate these factors, on the basis of Fuzzy Controller, the study function should be introduced. Meanwhile in order to establish the efficient algorithm of Self-Learning Fuzzy Controller, we make the improvement to the conventional fuzzy algorithm. Then according to the basic idea of self-learning control, we design the Self-Learning Fuzzy Controller.Nextly, the simulation study on the self-learning fuzzy algorithm applied to control systems is carried. The mathematical model of material testing system Established, compare performances of the system simulation, using the MATLAB software, with two different strategies of Fuzzy Control and Self-Learning Fuzzy Control.Finally, taking LabVIEW as the software developmental platform of virtual instrument, intelligent controllers are set up separately based on the fuzzy algorithm and self-learning fuzzy control algorithm, which are carried on the computer real-time control experiment to the material testing machine system.Theoretical analysis and experimental results indicate: it can effectively revise fuzzy control rules, and improve dynamic response characteristic of the system, after the Self-Learning Fuzzy Controller studing many times.