Fault Detection And Control Reconfiguration Based On Mixed Logic Dynamic Model

Hybrid system is refers to the continuous variable system (CVS) and discrete eventsystem (DES) interaction with each other to form a unified dynamic system. The rapiddevelopment of modern industrial technology greatly improve the production efficiency,atthe same time, the possibility of various failure also follow up. Once the system failure occurs,it may lead to the significant loss of personnel and property. In the process industry, thehybrid system characteristics is widespread. Therefore, the research of fault diagnosis andfault-tolerant control based on the hybrid system theory is of great significance. this paperhas researched the prediction control problems, fault diagnosis issues and the controllerreconstruction algorithm Based on hybrid logical model of the hybrid system. Take the three-tank water system as an example, we do a simulation study, the simulation results show theeffectiveness of the algorithm of this paper. The main research content and the resultsachieved are reflected in the following aspects:First of all,This paper has studied the control problem of hybrid system based on thehybrid system mixed logic dynamic model. On the basis of the establishment of mixed logicdynamic model,We have researched the optimal control, predictive tracking control andhybrid predictive control based on mixed logic dynamic model. These control optimizationproblems can eventually be converted into a standard mixed integer quadratic programmingproblem to solve. Through the simulation of hybrid system predicted control on the three-tankwater system proved that the hybrid predictive control to be able to handle the optimizationproblem of hybrid systems with a large number of constraints.Secondly, The fault diagnosis problem of hybrid system has been studied.We havedetect the state of the system with the moving horizon estimation,based on the establishmentof a mixed logic dynamic fault model.The moving horizon estimation problem is transformed into a mixed integer quadratic programming problems to solve. In this chapter we haverespectively researched actuator fault and sensor fault based on hybrid system. By theimplementation of the three-tank normally open, normally closed fault, sensor drift and slowdrift fault detection simulation study, has verify the validity of the fault detection method wediscussed in this article.Finally, on the basis of the hybrid system fault detection, the fault-tolerant control ofhybrid systems were also studied. This paper used the reconfigure the control rate based onthe active fault tolerant control method. This is a model predictive controller with fault-tolerant performance. It use normal system MPC control to control the system when it isnormal. When the system detects a fault occurs, the use of2norm performance indicators,and the infinite norm indicators, based on quadratic programming and linear programmingfault-tolerant the control method has been studied, by reconfiguring the model predictivecontroller such that the requirements of the system there is a fault condition remains normalrunning and can meet the performance indicators. By two-tank water system fault-tolerantcontrol study, indicating that the2-norm performance indicators and infinite norm indicatorscan achieve fault-tolerant control, infinite norm indicators has faster response speed.