Trajectory Tracking Control Of Omnidirectional Mobile Robot Based On Model Predictive Control Algorithm

Omnidirectional mobile robots are a kind of holonomic robots that can control translation and rotation independently.They stand out especially in complex and narrow space because of their high maneuverability.The trajectory tracking control of omnidirectional mobile robots that satisfies the input constraints has become a hot research topic in the mobile robot motion control.In the process of application for an omnidirectional mobile robot,both external disturbances(such as friction)and uncertain model parameters have a negative impact on control performance.Therefore,this thesis studies the high-precision and strong robust omnidirectional mobile robot trajectory tracking control system that satisfies the input constraints.The main research contents are as follows:Firstly,the experimental platform of the three-wheel omnidirectional mobile robot is designed and built.Then based on the force analysis of it,the dynamic model of the omnidirectional mobile robot that takes unknown friction in consideration is set up and the characteristics of the model are analyzed,which lays a foundation for the control algorithm design and simulation as well as experiment research.Secondly,aiming at friction disturbances and complex input constraints in the trajectory tracking control of the omnidirectional mobile robot,an improved extended state observer based model predictive control system is proposed.The control system is designed as a cascade control system,whose external loop is designed to generate the velocity references given to the model predictive controller according to the trajectory references and the robot inverse kinematics.The internal loop control system consists of two parts.One part is the improved extended state observer that is designed to estimate the unknown friction accurately while restraining the initial derivative peaking phenomenon.The other part is the model predictive controller that is applied to compensate the estimated friction and accomplish the velocity control while satisfying input constraints.Stability analysis shows that bounded-input bounded-output stability(BIBO)is guaranteed for the control system.Simulation and experiment results validate the effectiveness of the proposed control system.Thirdly,aiming at the uncertainties of model parameters in the trajectory tracking control of the omnidirectional mobile robot,an adaptive model predictive control system is proposed.The control system is designed as a cascade control system,whose external loop is also designed to generate the velocity references given to the model predictive controller according to the trajectory references and the robot inverse kinematics.As for the internal loop control system,the adaptive updating law of the uncertain parameters is designed firstly according to the minimal cost function of the control system based on the gradient method.Then the adaptive incremental model predictive controller is designed by using the estimated parameters to complete the trajectory tracking control of the robot while compensating the friction and satisfying input constraints,which improve the robustness of the control system.Besides,it is proved that the parameter estimation errors are bounded and the control system satisfies feasibility as well as bounded-input-bounded-output stability.Simulation and experiment results demonstrate the effectiveness of the proposed control system.