Model Predictive Control Of Nonholonomic Systems For Trajectory Tracking Under Complex Constraints

Nonholonomic system is suitable for trajectory tracking because its state depends on trajectory realization in physics and mathematics.For model predictive control,the predictive state of the trajectory is used for tracking control of the current input in nonholonomic systems.By solving an optimization problem,the optimal control input is obtained to ensure tracking performance.In this paper,trajectory tracking model predictive control for nonholonomic systems is studied.The main research work is shown as follows:Firstly,a self-triggered model predictive control strategy is proposed for trajectory tracking control of nonholonomic systems with adaptive transmission intervals under multiple constraints.The next triggering time and the adaptive transmission intervals are obtained to discrete a control sequence into multiple parts through a self-triggered mechanism.The application of the control sequence reduces computational burden by a sample-and-hold way.The terminal tracking error eventually converges to an adaptive terminal set.In addition,the simulation results show effectiveness of the proposed self-triggered model predictive control strategy.Then,a switched model predictive control strategy is designed for trajectory tracking control of nonholonomic systems with dual terminal sets and adaptive dwell-time.Cost functions are switched to improve performance by a switching signal with the adaptive dwell-time under multiple constraints.A dual-terminal set is used to adjust trade-off between computational complexity and control performance.In addition,simulation examples show effectiveness and superiority of switched model predictive control for nonholonomic systems.Finally,an event-triggered model predictive control is studied for tracking and formation of a multi-vehicle system with collision avoidance and obstacle avoidance.An event-triggered mechanism is established to reduce computational burden.A compatibility constraint is proposed to ensure collision avoidance and convergence of multi-vehicle system by limiting an uncertainty deviation of each vehicle.An obstacle avoidance constraint is used to ensure a safe distance between each vehicle and obstacles.In addition,a simulation example is given to illustrate effectiveness and advantages of the self-triggered model predictive control strategy.