An adaptive interactive multi-model navigation method based on UUV

Abstract

This paper introduces an adaptive interactive multi-model algorithm for the navigation system of the unmanned underwater vehicle (UUV). The underwater environment is diverse and time-varying. Obtaining high-precision navigation is very important for the safety and effectiveness of autonomous navigation of unmanned underwater vehicles, which makes the reliability of UUV navigation-related sensors particularly demanding. When the underwater environment has absorbing substances or a deep-sea environment, the Doppler velocity log (DVL) will produce outliers or even fail. When the vehicle navigates to a flat terrain area and terrain aided navigation (TAN) performs navigation parameter matching, multiple adaptation positions will appear, resulting in abnormal measurement data. Similarly, when UUV sails to a strong magnetic area, magnetic compass (MCP) will be disturbed, etc. To improve the robustness of navigation and reduce the influence of the change of system model and measurement noise model caused by environmental factors, an adaptive interactive multi-model algorithm is proposed, which is carried out with the UUV based INS/DVL/MCP/TAN integrated navigation system. For the problem that the measurement noise of the system changes, different models are organized into a collection, and according to different situations, the corresponding model or hybrid model is selected in the collection of models to describe the measurement noise, so as to realize the seamless switching of the system model. The simulation results show that the adaptive multi-model algorithm can effectively control the impact of system noise model uncertainty on navigation accuracy, so that UUV has higher accuracy and less computation than single navigation and traditional algorithms, and has robustness and scalability.