Event-triggered Model Predictive Control Strategy And Its Application Research

Model Predictive Control(MPC)has shown great potential in dealing with the control problems of complex constrained and multivariable systems.It has been successfully applied in many industrial fields and has gradually become one of the most common optimal control strategies in modern industrial environments.The traditional model predictive control method adopts the mode of periodic rolling optimization.At the same time,the optimization control problem is more complex,which makes the on-line calculation more heavy and limits its application scope in the actual control system.Especially,when the communication and computing resources of the system to be controlled are limited,it is more difficult to ensure good control performance.And event triggering mechanism of model predictive control strategy is used in balancing system resources and ensure the system can do better control performance,can significantly reduce the computational burden of the online optimization controller,for the model predictive control method applied in actual industrial system provides a new solution,has important theoretical value and practical significance.In this paper,the design scheme of robust model predictive control based on event trigger is proposed and applied to the linear discrete system with constraint and disturbance.The specific work is as follows:1.For linear discrete systems with system constraints and bounded external disturbances,the design of the two-mode robust model predictive controller is introduced,including the linear matrix inequality(LMI)method,the robust invariant set closed-loop min-max method,the robust stability analysis of the optimal control problem,and other simulation studies verify the effectiveness of the two-mode robust model predictive control2.For linear discrete systems with system constraints and bounded external disturbances,a model predictive control method with decreasing predictive step size is proposed.The closed-loop min-max method is used to establish the robust predictive control optimization problem of the system,and a new variable substitution is introduced to solve the non-convex problem of closed-loop min-max predictive control.Based on the control strategy of decreasing prediction step size,the number of predicted steps at each optimization moment decreases in turn.The influence of bounded disturbance is considered in the construction of terminal constraints and control input constraints.The proposed method not only ensures the robust control performance of the system,but also makes the system state enter the terminal constraint set within the specified predicted step size.3.For bounded perturbed linear discrete systems with system constraints,a scheme for designing self-triggering model predictive controller is proposed.Multiple pseudo-terminal sets are obtained by off-line calculation,and the performance of the system is optimized by using the optimization performance index and control law based on the nominal system state.On this basis,the system terminal constraint is added to make the system state enter each pseudo terminal set in order within the specified step size,and a self-triggering robust predictive controller is constructed.Finally,by applying the algorithm to numerical system and actual Boost DC-DC boost transformation control,it is verified that the self-triggered robust predictive controller can guarantee the robust control performance of the closed-loop system,and significantly reduce the average sampling frequency of the system and the on-line calculation of the controller.