The Research On The Application Of Neural Network Technique In The Inverted Pendulum System

The inverted pendulum is a typical nonlinear, strong-coupling, rapid, multivariable and absolutely instable system, whose control is a typical example in application of control theory, and usually it is used to test the validity of control strategy. Simultaneously, due to the similarity between the control of inverted pendulum and the control of rocket and aero craft, the control algorithms and conclusions can guide other projects of control.The research of the inverted pendulum system not only increase the stages, but the most important thing is to develop the existing control methods and study the new control methods, moreover apply new control methods to real systems, meanwhile propose new control tasks. The paper researches the control of the inverted pendulum by using intelligent control algorithms based on 1-stage and 2-stage inverted pendulum. The main works of the article are:1. Discussing the mathematics modeling methods on 1-stage and 2-stage inverted pendulum, and deducing their nonlinear differential equations as well as state equations after linearization.2. Discussing and combining GA and BP algorithm based on finding their advantages and disadvantages, and building a neural network controller, which can be used in linear 1-stage and 2-stage inverted pendulum of Googol Corporation, and proving the validity of this structure and learning algorithm.3. Using the gensim functions to produce the module and connecting it to the normalized module, building a neural network controller in the simulink toolbox and then taking an effective real-time control of the inverted pendulum.4. Analyzing the stability and the anti-jamming ability of neural network when it controls the inverted pendulum, and giving the simulated pictures, the real-time control curve and the real-time curve of 2-stage inverted pendulum under perturbation.The experimental results indicate that the neural network integrated GA and BP arithmetic is an effective control method in the control of inverted pendulum and it has a very good effect in the control of the inverted pendulum.