Research Of Information Fusion On INS/GPS Integrated Navigation Based On BP Neural Network

The INS/GPS integrated navigation system is a modern navigation system which integrates the independence of INS and the high degree of accuracy of GPS. It is the one of the developing directions of modern navigation system. Artlfieial Neural Network, a new but hot cross discipline, enjoys quite a great prospect of application. With its many characteristics such as large-scale parallel information processing, excellent self-adaptation and self-learning, the Network has been used in solving the complex nonlinear dynamic system.The thesis has a detailed discussion on INS and GPS, elaborating the parts of each system, principles of positioning, mistake model and model calculating, and accounting for the necessity of combing the INS and GPS.Then the paper introduces the basic theory and concepts of ANN. It focuses its research on the problems of back-propagation network structure and algorithms principle. Then it discusses some of the problems that lie in BP Neural Network. Then the paper applies it to the information fusion on INS/GPS integrated navigation, built a BP model, decides the hide codes and parameters such as momentum and learn rate, and so on. The results of application show that the integrated navigation system is of high accuracy for prediction and therefore meets with actual requirements of application.At last the track simulator, gyro emulator and accelerator emulator are established. Then use matlab to build the integrated navigation flat. Contrast with INS, the INS/GPS integrated navigation system enhance navigation precision and satisfy the navigation requiring.