The Balance Control Of Single Linear Inverted Pendulum Based On Dspace Simulation Platform

The inverted pendulum system is a naturally instable nonlinear system thattakes on over coupling, multivariable features. The simple structure makes it easy tochange parameters and display control effect. It is an ideal experimental platform totest the correctness of control theory and controller performance. The controltechnique of the inverted pendulum has great application value in the field ofrobotics, aerospace, military and industrial process control technology.This paper combines theoretical analysis on offline simulation andsemi-physical simulation experiments together. We use dSPACE simulation platformto help us adjust controller parameters online, watch on system states, and textcontroller performance quickly. The balance control of the inverted pendulumsystem is achieved with big noise in angle signal.Firstly, this paper establishs the mathematical model of system with theapplication Lagrangian equation of analytical mechanics, and gets the mathematicalrelation between key variables to analyze system characters. We apply fuzzy controlalgorithm for getting better robustness. A single input rules module dynamicallylinked fuzzy control(SIRMs)is raised, which can overcome weakness of generalfuzzy controller, such as excessive input variables, complex fuzzy rules, etc. SIRMscontroller combines advantages of state feedback control and fuzzy control, andsimplifies inference framework. Compared with Linear Quadratic Regulator, SIRMscontroller has better robustness and smooth control procedure.In the physical system, the angular velocity of the pendulum can not bemeasured directly. When angle of the rod is measured by potentiometer, the angledetection signal is accompanied by measure noise on over the frequency domain.Ifwe conduct difference operation on the swing angle signal to acquire angularvelocity signal, the noise will be enlarged. Introducing the two signals with noisesinto the controller directly causes the system to lose stability and damaged theequipment. Kalman filter is introduced to estimate the swing angle and angularvelocity signal as inputs of controller. KF filter cuts down the adverse effects ofnoise on the control system greatly.Finally, the real-time and semi-physical experiments are taken based ondSPACE simulation platform. The experimental results verify the effectiveness ofdesigned controller and KF filter.