On The Adaptive Control Algorithm And Experimentation Of Ball And Plate System

In this paper，trajectory tracking control for adaptive control algorithm isinvestigated on the ball and plate equipment BPVS-JLU II,which is independentlydeveloped by laboratory, and expect to be able to improve the accuracy of balltracking.The accurate mathematical model of the ball and plate system is difficult toestablish, as well as part of the control programs for the ball and plate system areapproximate linearization simplified model, thereby changing the inherentcharacteristics of the ball and plate system. Therefore, this paper uses the kineticanalysis and Lagrange equation to establish a complete mathematical model for theball and plate system,which is given a detailed introduction. First, the kinetic analysisfor the ball on the plate. Then, using the Lagrange equation method to derive themathematical model of the ball and plate system gradually. Finally, giving the statespace model of the ball and plate system. To further facilitate the implementation ofcontrol for the ball and plate system, based on some objective assumptions prerequisite,this paper also gives two typical simplified models of the ball and plate system, whichare the simplified model of ignoring the coupling between x-direction and y-directionand the simplified model of the comepletely linearization.The movement of the ball on the plate is more complicated, not only by theuncertainty of the system internal structure, and there are many external randomdisturbances. Therefore, how to ensure the control accuracy of the ball is veryimportant in such a complex environment. In this paper, exploiting adaptive controlcharacteristic of solving the control problem for the plant with structural uncertainty,parameter uncertainty, and the interference factors，combined with single neuronnetwork control in neural network control, combined with the traditional PID control, to design a single neurons network model reference adaptive PID controllersynthetically for the ball and plate system. This controller has a simple structure, asmall amount of calculation and real-time adaptive adjustment function for thecontroller parameter.The control and robustness for the designed neural network model referenceadaptive controller in the the ball and plate system, which is studied through simulationand experimental. First, establishing the simulation module of the ball and plate, thengrouping experiments for different conditions, a circular trajectory, the8-shapedtrajectory and variable trajectory tracking experiment and a comparative experimentfor the trajectory tracking of the ball in different running speed Mainly, each trajectorytracking experiment is added to the white noise or pulse interference, which objectiveexist in the actual system to simulate the uncertainty and disturbance. From the effectof simulation and experiment, the neural network model reference adaptive controllerhas a better tracking result and robustness.