The Research Of The Method Of Correction For Auv Navigation System At Different Depth

With increasingly competitive and further marine research work between countries around the world, The technology of underwater navigation of the Autonomous Underwater Vehicle which is the primary tool in marine development, is needed to obtain higher accuracy.Due to the complex environment of ocean, especially the deep ocean, errors of the AUV navigation system is inevitable. Therefore, the research on a simple and convenient method to correct the errors of underwater navigation is essential for AUV to operate the marine mission successfully.The purpose of this paper is to discuss a method to correct the errors of the dead reckoning system at different depth which will minimize the navigation error. The main work is done as follows:Firstly, when the depth of AUV is small, the DR/GPS integrated navigation system is designed to improve navigation accuracy. With the free switching between two operating modes, AUV can correct the errors of the dead reckoning system with GPS signals with high accuracy when it is available, so as to greatly reduce navigation errors. And it suggests that the integrated navigation system is effective by the sea trial.Secondly,the parameters fault-tolerance identification method for outliers eliminating,the five-spot triple smoothing algorithm and STUKF are used to reduce the noise contained in the data of navigation sensors. Through the comparisons of filtering result between STUKF and standard UKF, it is proved that the STUKF theory can improve real-time performance and the stability under abrupt catastrophe state with relatively accurate filtering result.Thirdly, the method of Range-Only positioning is discussed to get the initial position after the deep-diving of AUV. Then the result is used to correct the errors of the dead reckoning system, thus greatly reducing navigation errors. Meanwhile, the parameters which may affect the positioning accuracy are analyzed and the observability and stability of the Range-Only model are discussed.Finally, the function of fminsearch used in Range-Only positioning is theoretically analyzed and proved with simulations by MATLAB. What is more, simulations are carried out under different parameters which may affect the positioning accuracy by MATLAB,getting accurate positioning result. Therefore, the viability and feasibility of the Range-Only positioning method are verified.