Research On Robust Predictive Control Of Uncertain Systems

In the actual industrial production process,the system is easily affected by the external environment,which can result in irregular changes in its own structure and parameters.At this time,an accurate mathematical model adopted may not be able to describe the system accurately,which will lead to direct model errors between the actual system and its model.Therefore,the model errors must be properly handled to reduce or eliminate its impact on the modelling of the system.Because robust predictive control has a good ability to deal with model uncertainty and constraints,it will play a vital role in the face of the system with model uncertainty.Although the research on robust predictive control has achieved fruitful results and has been widely used in many fields,there are still some problems to be solved,such as how to ensure system performance of uncertain system with quantization and time delay.To the end,this thesis further studies the problem of robust predictive control for uncertain systems.The main research contents are as follows:(1)A robust predictive control algorithm is proposed for a class of uncertain systems with input quantization and time delay.Firstly,considering that the input signal of the system is quantized by a logarithmic quantifier,a model of the norm bounded uncertain system with input quantization and time delay is established by using the sector bounded method.Then,according to the established system model,the stability problem of the uncertain system is studied,and the sufficient conditions of the stability of the system are given in the form of a set of linear matrix inequalities,and a robust predictive controller that can make the system stable is designed.Finally,the effectiveness of the algorithm is verified by comparing the simulation results under different quantization densities.(2)A distributed robust predictive control algorithm is proposed for uncertain systems with multiple subsystems.Firstly,considering that each subsystem has information exchange and influence with other subsystems.Without loss of generality,it is assumed that the state of each subsystem is affected by the input and state of the other subsystems,and then a subsystem model with norm bounded uncertainty can be established.Furthermore,a distributed robust predictive controller is designed according to the subsystem model,and the sufficient conditions for the stability of the subsystem are given.Finally,the effectiveness of the algorithm is verified by data simulation.