| In the modern society where the network information technology develops rapidly,the networked control systems(Networked control systems,NCSs)has been applied in a wide range of industrial control fields.However,it also brings some challenges in practical applications.In the network environment,the transmission of data needs to consider the situation of quantification and packet loss,but the traditional predictive control approach will not consider these factors.If these factors are ignored,it will inevitably cause some inestimable actual industrial production loss.In addition,some disturbance factors in the actual system also need to be taken into account.In summary,according to the characteristics of model predictive control(MPC)methods,this paper considers the quantized error and packet loss in the network control system with bounded disturbance.By utilizing the interval type-2(IT2)Takagi-Sugeno(T-S)fuzzy model and Hammerstein model describe the nonlinear of the networked system.The quadratic bounded technique and linear matrix inequality(LMI)technique is used to design and solve the optimization problem.The main research content of this article is divided into the following three parts:1.Considering the nonlinear NCSs with quantization,packet loss,and bounded disturbance,a method of state feedback predictive controller based on IT2 T-S fuzzy model is proposed.Firstly,an event-triggered strategy was applied to reduce the burden in the network environment,and the quantization and packet loss that occur in the network are taken into count to establish the closed-loop system.The phenomenon of packet loss is described as a Markov chain model,and the quantization error is processed by the sector bound approach.For the bounded disturbance that existing in the system,which can be disposed by the quadratic bounded method.Finally,the MPC method is used to solve an optimization problem at each sample time,and the first of the optimal sequence of the solution is applied to the system to achieve the stability control of the closed-loop system.The recursive feasibility and random stability of the established control method are proved.2.For the IT2 T-S fuzzy NCSs with random bounded disturbances and packet loss,an event-triggered multi-step output feedback predictive control approach is proposed.The packet loss phenomenon of the network environment is described as Bernoulli binomial distribution,furthermore,an event triggered mechanism is used to decide whether the data should be released into network or not,which can save network resources effectively.Then,by introducing the quadratic boundedness technique,the sufficient conditions for describing the closed-loop stability of the networked control system under the bounded interference are obtained.Based on this stability condition,a multi-step output feedback predictive control method under network environment is proposed.And on the basis of this algorithm,an auxiliary optimization problem is added to reduce the boundary of state error and improve control performance.Finally,the recursive feasibility of the multi-step output feedback predictive control method is analyzed,and the stability of the closed-loop system is proved.3.For the nonlinear NCSs with random bounded disturbance,packet loss,and quantization error,a multi-step output-feedback predictive control method based on the Hammerstein model is proposed.Firstly,considering the phenomenon of data loss and quantization error that exists in the systems,which can be appropriately disposed by the Bernoulli data loss model and sector bound approach,respectively.Then,based on the quadratic boundedness technique,the synthesis approach of multi-step output-feedback predictive control method can be obtained by online solving the optimization problem of minimizing the upper bound of the quadratic performance function.Finally,in order to further improved the control performance,an auxiliary optimization problem is added to tighten the bound of state error. |
PDF Full Text Request