Studies On Fault Diagnostic Technique Of Integrated Navigation System Based On Neural Network

With the rapid development of electronic computer and the progress of modern control system theory, especially the development of aviation, spaceflight and navigation, people have a high request to navigation ability, automation degree, reliability and precision of integrated navigation system. So the traditional fault-diagnosis technique has not already satisfied the need of the modern integrated navigation system, therefore must adopt the intelligence fault-diagnosis technique. The artificial neural network that is quickly developing and extensive applying to the field of fault diagnosis develops a new approach for the fault diagnosis of integrated navigation system.In this paper, with the SINS/GPS/DVL integrated navigation system for research object, from the information fusion technique of integrated navigation system, neural network technique is used to information fusion and fault diagnosis of the integrated navigation system. The paper puts forward the design method of neural network for the fault diagnosis, finally designs an intelligent fault-tolerance integrated navigation system, which have a high precision and reliability of integrated navigation system.Firstly, this paper carries through overview of fault diagnosis technique and analyzes the research actuality of the fault detection and diagnosis in the integrated navigation system on the basis of the reading and analysis of the domestic and international concerning datum. And, it introduces the relevant knowledge of the neural network, and carry on thorough theories research to neural network especially BP neural network and combined mode of the neural network and fault diagnosis.Then, the paper introduces briefly information fusion technique which is one of the key techniques in the integrated navigation system. The information fusion technique includes kalman filter and federal kalman filter. On the basis of analyzing state equations and observation equations of the SINS/GPS/DVL...