Distributed Model Predictive Control With Output Feedback For Nonlinear Systems

Distributed model predictive control is an effective method to solve the control of largescale systems.It can achieve the control performance as good as possible under the circumstance of the communication in the system as simple as possible and the burden of communication as small as possible.Meanwhile,it can guarantee the convergence of the algorithm and the stability of the system.Because of the actual process usually tend to nonlinear,and the states of the system are not easy to measure accurately due to constraints,even immeasurable,state feedback is often difficult to achieve the desired aim in this case.As a result,output feedback control is taken into account,which is to estimate the states of the system through designing the state observer.For a class of nonlinear systems whose states are immeasurable and a class of nonlinear systems subject to delayed measurements and communication disruptions,respectively,a kind of output feedback distributed model predictive control is proposed.It is guaranteed that the estimated states of the observer are ultimately bounded and the error of estimated states and the actual system's states is bounded.As a result,the states of the actual system are ultimately bounded.The main job of this paper includes the following two parts:Firstly,the output feedback distributed model predictive control problem is investigated for a class of nonlinear systems whose states are immeasurable.First of all,a state observer is designed and it is proved the error of estimated states and the actual system's states is bounded in the case that the outputs of the system are sampled of asynchronous measurements.Then a Lyapunov-based controller is designed to make sure the asymptotic stability of the nominal observer.At last,an output feedback distributed model predictive control algorithm is proposed.It is proved that the estimated states of the observer are ultimately bounded,and then the states of the actual system are ultimately bounded.Secondly,the output feedback distributed model predictive control problem is investigated for a class of nonlinear systems subject to delayed measurements and communication disruptions.First of all,the model of the nonlinear systems is given,and the modeling of delayed measurements and communication disruptions are described.The state observer with time-delay is designed to prove the error of estimated states and the actual system's states is bounded.Then a Lyapunov-based controller is designed to make sure the asymptotic stability of the nominal observer.At last,an output feedback distributed model predictive control algorithm subject to delayed measurements and communication disruptions is proposed.In order to handle communication disruptions,a constrained feasibility problem is incorporated in the DMPC architecture.It is used to determine if the data transmitted through the communication channel is accepted or not.And it is proved the estimated states of the observer are ultimately bounded,and then the states of the actual system are ultimately bounded.