Research On Intelligent Control Of The Inverted Pendulum System

Based on math tool such as linear algebra and optimal method, modern control theory is able to resolve more complex systems than classical control, but it still can not meet the needs of modern technology, so it is required the introduction of a new control strategy. With the progressive development of the field of artificial intelligence, intelligent control in which to solve complex control object problems increasingly shows its advantages, and it shows a strong vitality in practical applications. The theory of intelligent control only has a few decades'history, but its existing application results and theoretical development describes intelligent control becoming one of the forefront subjects of automatic control, and it reflects one of the current development directions.Inverted pendulum system is full of multivariable, high-order, nonlinear, unstable characteristics, and its control process can effectively reflect many significant control problems, such as: stabilization problem, nonlinear problem and robust problem. So usually inverted pendulum system is used to testify the effectiveness of control theory. Research on it has not only meaningful theory worth, but also important engineering background and real significance.For the intelligent control of the inverted pendulum systems, the following content will be discussed in this paper:First, this paper makes a study of the constitution and work principle of the inverted pendulum system, re-computes the math model, and analyses the robustness and main disturbance sources.Second, study and research intelligent control neural network BP algorithm system, design a rule-based neural network controller, take advantage of Simulink simulation platform to connect the inverted pendulum model, and simulate successfully. Have good robustness on the level of actual control of inverted pendulum.Third, integrate fuzzy control and neural network by using intelligent control method, train membership function of the fuzzy controller by using learning ability of neural network, design fuzzy neural network controller, and improve learning ability of the whole system.Finally, after the completion of simulation on computer based on Simulink simulation platform, use MATLAB as control interface software. The researching results of this paper are applied to linear inverted pendulum experiment equipment produced by Googol Technology LTD. so as to verify the effectiveness of the control laws. During the experiments, the control parameters are adjusted and how the parameters effect on the performance is concluded. The results of experiments prove that each of the controllers has a good control effect.