Research On Markov Jumping System And Time-delay System With Disturbances For Fault Diagnosis

The goal of fault detection and isolation is to identify whether the system operation state is abnormal. First we should detect the system faults currently occurred, and then isolate the faults via some techniques. It can guarantees that the faults would not cause system fluctuations. This paper researches the system fault diagnosis based on the space geometry theory, which transform the system characteristics into space geometry characteristics and then use the space projective operation instead of the complex matrix calculation. Specifically, the method constructs the residual set related to fault signals by using the unobservable subspace concept in space geometry theory, and then makes the relationships between residuals and fault signals are corresponding to each other. Hence, the system faults detection and isolation are accomplished.The article research the problem of fault detection and isolation by using the space geometry theory in Markov jump system and time-delay system. Specifically, this paper establishes the appropriate fault model in different system, and designs the corresponding residual generator; Then to solve the corresponding residual generator parameters, accurate residual generator and its residual value are obtained; the relation between residual and fault could realize residual is sensitive to specific faults, and is not sensitive to other fault, which in the case of the associated with input disturbance, can complete the Markov jump system actuator fault and sensor fault or time-delay system sensor fault detection and isolation. On this basis, the article mainly studies on the following three situations:1) To solve the problem of actuator fault problems in Markov jump system with input disturbances. The continuous-time Markov jump system with actuator fault is molded and the residual generator is designed. Specifically, for the case of system jumping, we use H?optimization technique to reduce the effects of disturbance inputs on the residuals based on the unobservability subspace and space division projective operation of space geometry approach, and then the corresponding filter parameters are obtained based on the space geometry approach, so that the residual signal and actuator fault satisfy correspondence, and through the residual output value compared with the preset threshold to determine faults in the system, finally complete the actuator fault detection and isolation in Markov jump system.2) For the sensor fault problems in Markov jump systems with uncertain disturbances. The sensor faults can be translated into the state inputs using the measurement equation. Subsequently, a cluster of residual generators is designed by employing the space geometric method. The corresponding filter parameters are obtained based on the space geometric approach, then using H?optimization technique reduce the effects of disturbance inputs on the residuals, at the same time, the residual generator is designed so that the residual signals and sensor faults satisfy correspondence, which can be used to detect and isolate the sensor faults in Markov jump systems with disturbance.3) For the sensor fault problems in time-delay systems with uncertain disturbances. Using the measurement equation, the sensor faults can be translated into the state inputs. Subsequently, a cluster of residual generators is designed by employing the space geometric method. The corresponding filter parameters are obtained based on the space geometric approach, then using H ? optimization technique reduce the effects of disturbance inputs on the residuals, at the same time, the residual generator is designed so that the residual signals and sensor faults satisfy one to one correspondence, which can be used to detect and isolate the sensor faults in time-delay systems with disturbance.