Distributed Predictive Control Of Several Types Of Nonlinear Systems

For complex industrial system features such as nonlinear,coupling of multiple subsystems,distributed model predictive control?DMPC?is an effective way to solve such problems.DMPC's every subsystems in structure is connected with each other through network,information sharing,DMPC can able to work on local control and relatively independent.For switched nonlinear systems,due to internal and external factors of the system state is measured,the third chapter is presented in this paper a state observer,the output feedback to prove the real state of the system and estimate the state of the value is bounded;Then using a Lyapunov function controller design for a distributed controller was given,₁ controller can guarantee the stability of the system,the controller₂ with controller₁exchanges and cooperation to improve the overall performance of the system.After designing a distributed predictive control algorithm based on output feedback,the stability of system is proved.However,when reached the desired performance,the system will cycle for predictive control,this will waste system of online computation time;After when the system reaches the expected control effect,the system will still be periodically for predictive control,this will waste system of online computation time,in this paper,the fourth chapter has designed a distributed predictive control algorithm based on event trigger,when events trigger condition is met,predictive control and save the control system,when the system does not satisfy the trigger condition,will use a trigger point to save on the control effect on the system,and combining the characteristics of the input state stability to prove the convergence of the system.The work of this paper can be expanded into the following two parts:First,the conventional large-scale high-latitude systems usually adopt the centralized predictive control,however because of the centralized predictive control on the amount of calculation and performance indicators already cannot satisfy the requirement of modern industry.This paper adopts industrial common switched nonlinear systems,designed a distributed model predictive control algorithm based on output feedback.In the case of system status,unpredictable,proved by output feedback system bounded real state and estimate the state of difference,and then use a given controller based on Lyapunov function,a distributed predictive controller was designed,and presents a distributed predictive control algorithm,the stability of system is proved.Second,for a class of discrete-time nonlinear systems,when there is a delay the correlation information between system this paper designed a is only related to the subsystem of the event trigger module,and gives the event trigger condition.When events trigger condition is met,the subsystem for predictive control and save the current moment control;When the trigger condition is not met,the subsystem will use a trigger point save control effect on the system,then presents a distributed predictive control based on event trigger control algorithm,combined with the input state stability,the convergence is proved.