Research On AUV Multi-source Integrated Navigation Data Fusion Method

Underwater navigation technology is not only the technical support for autonomous underwater vehicle(AUV)to perform its mission,but also the prerequisite to ensure its safe return.Integrated navigation technology based on multi-source sensor data fusion is the future development direction of underwater navigation.Aiming at the problem that the statistical characteristics of measurement noise of AUV navigation system are unknown or changed due to the complex and changeable underwater environment,this paper studies the data fusion method of multi-source integrated navigation of AUV.Firstly,the AUV multi-source integrated navigation system is designed and the filtering algorithm is studied.The multi-source integrated navigation system which takes Strapdown Inertial System(SINS)as its main navigation system,while Ultra Short Baseline Positioning System(USBL),Doppler Velocity Log(DVL),and Magnetic Compass(MCP)as its auxiliary navigation system are designed.Studiying the filtering performance of three linear filtering algorithms under time-varying measurement noise,and the simulation experiment is carried out to verify the performance.Secondly,the error model of navigation sensor is established and the strapdown inertial navigation simulation platform is designed.Based on the introduction of some commonly used coordinate systems and coordinate transformations for navigation solutions,the principle of SINS,USBL,DVL,MCP are deduced and analyzed,and the error models are established.And the simulation platform of strapdown inertial navigation system is designed.The motion trajectory can be simulated by inputting AUV motion parameters,and the corresponding measurement data can be obtained according to the working principle of the sensor.The simulation experiment is carried out to verify the effectiveness of the platform.Then,the integrated navigation system model is established and the Sage-Husa adaptive federated filtering algorithm is studied.Based on the introduction of federated filter,the system equations and measurement equations of SINS/USBL,SINS/DVL and SINS/MCP are deduced.And aiming at the problems of Sage-Husa adaptive filtering algorithm,an improved Sage-Husa adaptive federated filtering algorithm is proposed when the measurement noise is unknown or changing.Finally,interactive multi-model adaptive federated filter algorithm is studied.The interacting multiple model theory is introduced into the improved Sage-Husa adaptive filtering algorithm,and according to its estimated variables measuring noise to set up a set of dynamic model to approximate real system,an interactive multi-model adaptive federated filter algorithm is proposed to improve the accuracy of Sage-Husa adaptive federated filtering algorithm,and designing simulation experiment to verify the validity of the proposed algorithm.