Research On Realtime Control For Rotary Inverted Pendulum

Inverted pendulum system is a complex, multivariable, nonlinear, strong-coupling,unstable and high order system. Among the installations and equipments of invertedpendulum, rotary inverted pendulum device is a typical one. Compared with straight-lineinverted pendulum, though rotary inverted pendulum is more simple in structure, while itsmodel is more complicated and unstable. As a result, it makes control harder. Invertedpendulum can be controlled successfully in many different ways. And we can test and verifythe correctness and practicability of a control algorithm, which is proposed recently, by meansof inverted pendulum device. Therefore, the study of inverted pendulum has a profoundsignificance in checking and improving control algorithm.In this paper, the energy of pendulums are analysed by applying Konig theorem.Computing results prove that the proposed analytic method is perfectly right and it is betterthan traditional analytic approach for reducing the computing burden. Then, the mathematicmodel of both rotary double inverted pendulum system and rotary single inverted pendulumsystem can be deduced by applying Lagrange method. On this basis, the researches onbalancing control of an inverted pendulum system are as follows: realtime control andsimulation on a rotary single inverted pendulum system in generalized predictive control andparallel control of double fuzzy controllers; realtime control and simulation on a rotary doubleinverted pendulum system in LQR, fusion function algorithm based on LQR, ANFIS(adaptive neuro-fuzzy inference system), state feedback control based on T-S fuzzy model andgradual fuzzy control based on T-S model. Simulation on a rotary double inverted pendulumsystem in generalized predictive control, parallel control of triple fuzzy controllers and BPneural network principle.The results of realtime control and simulation show that the proposed control approacheswork excellently in balancing control of an inverted pendulum system. Especially, not onlycan the gradual fuzzy control based on T-S model ensure the stability of the system, but also itcan improve the control precision. The effectiveness and superiority of this algorithm areshown by comparing realtime control results with other control methods.