Robust Fault Tolerant Control Method Based On Neural Network

The system has high security and reliability means that staff personal safety can not only be better protected, but can also improve economic efficiency and productivity. Fault Tolerant Control (Fault Tolerant Control, the FTC) aims at maintaining the stability of the system in case of failure, where the system can automatically compensate for failure before the performance deterioration as much as possible so as to restore the system failure to ensure the stability and reliability of the system. Therefore, the fault-tolerant control theory has opened up new ways to improve the reliability and security of the system. Dynamic system fault detection has thus been used as an independent branch of classical control theory (Fault Detection and Diagnosis, FDD) and technology which become an important support for fault-tolerant control technology. Fault detection task is to describe the characteristics of the system failure, and use it to find out unexpected failure. In order to achieve good fault tolerance effect, there is an urgent need for efficient online fault detection mechanism to provide accurate fault information so as to complete the system’s fault-tolerant control functions.The necessary condition of a system to be fault-tolerant means that there is the redundancy of the accessories, such as the actuator tolerant actuator drive redundancy, where the measurement redundancy of the sensor for the sensor fault tolerance exists. The key to the fault-tolerant control system design is to achieve the purpose of redundant use of these devices. Then the way to deal with failure and redundancy in the system can be composed. Fault-tolerant control system can be divided into active fault tolerant systems and passive fault-tolerant system. Active fault tolerant control system uses an adaptive online identification of fault parameters, or fault detection mechanism for fault information. Passive fault tolerant control system design does not require fault information online, where system failure can occur online in the control system design as a priori knowledge. They use the robust fault tolerant control technology. Active fault tolerant control and passive fault tolerant control depend on the system model:namely qualitative model and semi-quantitative or quantitative, etc.The actuator is the most prone to failure due to the implementation of the long uninterrupted control tasks. Tolerant control strategies for actuator fault-tolerant control system design are the most important aspects to be considered. Fault-tolerant control originated in the aircraft (such as the field of aviation, aerospace) has entered the industrial process control through continuous promotion and application areas. The areas of fault-tolerant control system design have therefore a very important significance. Tolerant control strategies that address mostly to the failure of the actuator will be simply discarded, thereby ignoring actuator failures that may exist in the remaining driving force, the fault tolerant control system can be obtained using fault-free actuators. Obviously, this is very unwise to reduce the overall drive capability of the system, and reduce the fault-tolerant control strategy design freedom. To this end, the failed drive in fault-tolerant control should be considered. If possible after a fault it should be all of the driving force in the system all after a reasonable re-organization for fault-tolerant control system.For system fault detection and fault-tolerant design of the control system, and conducted in-depth research and discussion on hot issues on fault-tolerant control, the main content of this dissertation and the research work are listed as follows:1. The background and significance of the paper is described, the development of fault-tolerant control has been pointed out together with the hot topics of the fault-tolerant control research.2. The robust fault tolerant control problem for linear uncertain continuous system has been developed. Using linear matrix inequalities (LMI), sensors and actuators, component failure situation, the full system robust fault tolerant control conditions, the results presented in the form of linear matrix inequalities. Implementation of a simulation study has been carried out.3. Delay uncertain continuous system state feedback robust stabilization problem has been presented. Based on the theory of linear matrix inequalities (LMI) and Lyapunov stability theory, a sufficient condition has been established for the system asymptotically stable in the case of sensor and actuator failures, and corresponding state feedback robust fault-tolerant control law design method has been established.