Research On Control Algorithm Based On Inverted Pendulum System

Inverted pendulum is multivariable, nonlinear, strong-coupling and instabilitynaturally. Also its controller can reflect reliability and effectiveness of control theory.Under conditions of understanding inverted pendulum's movement rules and model,controller is designed by using the known theoryto make pendulum stayupward that iscalled stabilization. This process helps us master the basic ideas of automatic controlsystem.In this paper, controllers are based on pole-place, LQR, fuzzy logic system,variable-structure control and energy management. On the premise of pendulum staystability, major task is to analyze advantages and characteristics between algorisms,also useing these to help other theories to deal with problems that they face. It reducesthe input variable dimension of fuzzy logic controller by designing a fusion functionapllying optimization control theory, and solves the problem of "rule explosion"successfully. Using learning capability of nerve network to train membership functionand fuzzy rules in Adaptive Nero-Fuzzy Inference System (ANFIS), fuzzy controllerof inverted pendulum is established and by this way it solves the problem that fuzzyrules obtain difficulty. What's more, it optimizes parameters of fuzzy controller andpromotes its performance. Finally it realizes inverted pendulum system's practicalitycontrol based on control algorithms through Simulink/RTW, and obtains satisfyingeffects. The results indicate that combination of different algorithms can reduce thecontroller's complexity, improve the performance, also be a new thought to designsuitable controller. The computer simulations and real time control on the system areperformed and show that all above theories can manage stability control of invertedpendulum efficiently.All in all, combination of theories can be looked as a trend of controldevelopment. More factors be considered, best controller be designed.