Mobile robot control using self triggered model predictive control and PID theory

DOI:	10.11809/bqzbgcxb2024.11.030
Keywords:	mobile robot; model predictive control; self triggered model predictive control; PID control; dual mode control
Abstract:	Nonlinear mobile robot has become an indispensable intelligent device, thus it is very important to research for its control algorithm.Since self triggered model predictive control (STMPC) can not only be indicated to deal with robot constraints but also reduce resource consumption, the application of STMPC in robotics has been widely concerned.However, the existing STMPC strategy constructs the self triggering condition only by the error of a single sampling moment, ignoring the state mutation of the system due to the accumulation of state error.Therefore, in order to solve the problem of system state mutation due to additional perturbation in the traditional self triggered model predictive control for a mobile robot system, a proportional, differential and integral (PID) based STMPC (PID STMPC) algorithm is proposed.Firstly, based on the two wheel differential mobile robot system model and its additional constraints, a “Dual mode model predictive control (MPC)” controller is designed; secondly, a self triggering mechanism based on PID of error information between the actual state and its optimal prediction is constructed by ensuring that the cost function is strictly decreasing, which not only reduces the solving frequency of the optimal control problem (OCP) but also weakens the effects of residual error and state mutation on the robot system; then, the PID STMPC algorithm is designed by combining the “Dual mode MPC” and the self triggering mechanism based on PID information; finally, the effectiveness of the developed PID STMPC algorithm is verified by experiments in a real mobile robot system.As shown in the experiment results, compared with the standard STMPC strategy, the proposed PID STMPC method can not only effectively improve the control performance, but also reduce by 84.62% computational resources while shortening the response time by 54.63%.
Published:	2024-11-30
PDF