二级倒立摆系统的模糊控制策略研究-控制理论与控制工程专业论文.docxVIP

Abstract Inverted pendulum, which is a dynamic system with nonlinearity, multivariables, high equation orders, strong coupling, fast reaction and instinct instability, and has abundant and complicate dynamic behaviors, has become an ideal object and typical experimental equipment for verifying a certain kind of control theory and technique. Several fuzzy control strategies are employed to solve the contradiction between the control performance and the number of fuzzy rule, and to effectively control the double inverted pendulum in this thesis. Firstly, the fundamentals, characteristics and progress of fuzzy control are discussed, and the present status of research on inverted pendulum is summarized in the thesis. The nonlinear mechanism model of the double inverted pendulum is built up by means of Lagrange equations, and its linearization and performance analysis of stability, controllability and observeablity are carried out. Secondly, modern control theory, fuzzy control, neural network (NN) as well as their mutual combination are systematically discussed, and several fuzzy control strategies are given. Information fusion based fuzzy control strategy reduces the dimension of its input vector and solve the problem of rule number’s explosion by combining the linear system theory with information fusion technique. Takagi-Sugeno based gradual fuzzy control strategy, which has fewer and more effective rules and can greatly simplify the design process of fuzzy controller, is adopted to stabilize the double inverted pendulum. Adaptive neural network-based fuzzy control strategy uses learning ability of neural network to train membership function and parameters of the fuzzy controller in Takagi-Sugeno model, and ANFIS (adaptive neural network-based fuzzy inference system) to model and control the complex inverted pendulum. Finally, simulation results in Matlab/Simulink indicate that the fuzzy control strategies proposed in the thesis not only can solve the problem of rule