An online learning-based fuzzy control method for vibration control of smart solar panel

Abstract

Vibration control problem of solar panel structure is an ultra-important problem which was studied by a lot of researchers. This research is devoted to design an “intelligent” control algorithm for this problem, which can learn autonomously without supervision or a priori training data. Based on this idea, an online learning-based fuzzy control method is proposed in this article. The online learning-based fuzzy control is composed of reinforce learning algorithm and fuzzy inference system. The learning algorithm learns the fuzzy rule base by interaction with the plant and changes rule base generated by policy via evaluative reward signal to realize the learning goal. In order to verify the presented control method, vibration control of a typical single-panel smart solar panel structure bonded with piezoelectric actuators is considered in this article. First, the dynamic model of the smart solar panel structure is established using finite element method. Then, the state-space equation of the control system is presented. After that, technique details of the online learning-based fuzzy control method are described. Finally, experimental validation is conducted. The experimental results show that the online learning-based fuzzy control method presented in this article can suppress the vibration of the smart solar panel effectively and faster than fuzzy control method.