Dissertation
Dissertation > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > Automation systems > Automatic control,automatic control system > Fuzzy control, fuzzy control system

The Application of Fuzzy Control and Neural Network in Planar Double Inverted Pendulum

Author HeFeng
Tutor WangTong
School Harbin Institute of Technology
Course Control Theory and Control Engineering
Keywords Planar inverted pendulum Fuzzy control Neural network Sliding mode control Reference compensation technique
CLC TP273.4
Type Master's thesis
Year 2008
Downloads 276
Quotes 0
Download Dissertation

With the development of research activity in control technology, research on the accurate control strategy for inverted pendulum is playing an increasingly important role in the complicated object control of industrial production. The inverted pendulum is widely used as a method of verification experiment platform for its multivariable,nonlinear and strong-coupling. Inverted pendulum mainly includes linear inverted pendulum、rotary inverted pendulum and planar inverted pendulum three categories,of which planar inverted pendulum belongs to a relatively new kind,comparing with the other kinds of inverted pendulum, the control of planar inverted pendulum is more challenging for the two-dimensional movement of the cart.As the structure of planar inverted pendulum is relatively complex,the results of research in the field are relatively few. This paper makes a comparative study upon the application of planar double inverted pendulum in the real system with two widespread intelligent control methods——fuzzy control and neural network.In this paper,the GPIP2002 planar double inverted pendulum of Googol Technology LTD is considered as the plant. The main work of this thesis is outlined as follows.First, the GPIP2002 planar double inverted pendulum was introduced. And the nonlinear model was established by dynamic method, and then, two decoupled linear models were obtained by linearization around the balance states.Second, the fuzzy sliding mode control (FSMC) and the neural network model reference compensation control were proposed in this research. For the complexity of the system and the influence of measuring error and friction, the deviation of the linearized model is a little big. The effect of conventional controller based on linearized model is inaccessible to be ideal.By contrast, the analysis of the experimental result indicates that both the fuzzy sliding mode control (FSMC) and the neural network model reference compensation control can compensate for the uncertainty of the model well.Third, a control software was developed for Googol GPIP2002 planar double inverted pendulum,which greatly facilitates comparison and analysis of these different algorithms.It not only achieves the optimal controller, but also the above-mentioned two controllers. The data collected with this software showed the effectiveness of the above-mentioned two controllers.

Related Dissertations
More Dissertations