| With the rapid development of information technology,we have stepped into the era of Industry 4.0 featuring digitalization,networking,and intelligence.As the general framework of a new intelligent information control system,NCS sets the controllers,actuators,sensors and network media at an organic whole.It is a closed-loop control system for data transmission and information exchange through a shared network.Such systems are not only more complex in structure,more spatially distributed,but also more sophisticated in terms of control performance.Therefore they face more and more problems and challenges,in which security and reliability issues have become a primary service.And it is an effective way to solve the problems NCS faced by designing the fault-tolerant control of the NCS to improve the safety and reliability.In view of this,this paper starts with severity differentiation of the actuator faults under the Discrete Event-Triggered Communication Scheme(DETCS),and fuses the idea of active and passive fault-tolerant control.Then the co-design issues between NCS fault tolerance control and network communications of non-uniform transmission under DETCS have been studied by designing better performance and fewer switching selection strategies.The specific work is as follows:1)Constructing the framework of NCS active and passive fusion fault-tolerant control with fault severity distinction.Taking the uncertain NCS as the research object,the DETCS is introduced into the NCS system.And the state feedback control strategy and the concept of“partial order”are applied to distinguish the severity of faults.The passive fault-tolerant controller is designed according to different fault levels on the offline,and the corresponding fault-tolerant control law is actively selected on-line according to different operating modes.The framework of active-passive fusion fault-tolerant control was constructed,which makes it possible for the co-design between NCS active/passive fusion fault-tolerant control and network communications.2)Presenting the offline design method of NCS active/passive fusion robust fault-tolerant controllers and communication trigger weight matrix under DETCS and designing the online switching strategy.Based on the established model,firstly,by constructing Lyapunov-Krasovskii functions and using improved Wirtinger inequalities,reciprocal convexity Lemma and other techniques,the synergetic calculating method of offline design with less conservative robust H_?and?-stability robust H_?fault-tolerant controllers and communication trigger weight sets is presented for four types of different severity fault pattern sets,such as"normal","minor","moderate",and"heavy".Secondly,the selection strategies of less switching and better performance was designed online,which make up for time-consuming defects of the traditional active fault-tolerant diagnosis and reconstruction and improve the system's fault-tolerant performance on the premise of ensuring the system performance of different fault levels.3)Realizing the NCS active and passive fusion fault-tolerant control of non-uniform transmission based on SAE's classification of actuator failure severity.SAE as one of deep learning technology,this paper introduces it into the system.Through the training and studying of effective information mining and feature extraction for different severity of actuator data,the SAE module with high precision classification ability is obtained.Then the trained SAE module is embedded into the NCS active-passive fusion framework.In addition,the severity type of the actuator fault is identified by SAE online and then the optimal robust fault-tolerant controller solved on the previous point is switched.Lastly,the simulation results show that this method can not only accurately and reliably determine the type of actuator failure,but also effectively solve the problem of system fault tolerance failure caused by the time-consuming for diagnosis and reconstruction. |
PDF Full Text Request