Research And Implementation Of Stability Control Methods Of Linear Inverted Pendulum

Inverted pendulum system, which is unstable, nonlinear, multivariable and strongcoupling, is also a simple laboratory equipment and an ideal experimental platform inthe control field. It could be used to verify the feasibility of new control methods andto make a comparison between the existing control methods. So it is widely used inmilitary, aerospace technology, robot control and general industrial process withsignificant engineering background.Control algorithms of linear double inverted pendulum, which is excogitated byGoogol Technology Ltd, are analyzed and implemented in this paper. And thesimulation and real-time control platform is developed with C++Language in VC6.0application under Windows2000system. The main content of this paper is listed asfollows:Firstly, the nonlinear mathematical model of linear double inverted pendulumsystem is built with Lagrange equations. And based on this model, the linear andnonlinear models are established by using S-function of Matlab. Secondly, the LQRcontroller optimized by the PSO algorithm, the BP and the RBF neural networkcontroller are studied and desired effects are achieved in both simulation and real-timecontrol. The experiment result shows that LQR controller is robust enough to linearmodel of double inverted pendulum, and the neural network controller could reducethe car drift after the system is stable. Finally, as the study of inverted pendulumdepends on Matlab, which limits the openness of the application, a full simulation ofLQR algorithm based on VC6.0is developed. And the real-time control under thissoftware platform is elementarily passed with real-time performance to be improved.