Research On Underwater Navigation Method Based On Kalman Filter

Underwater vehicles play a crucial role in exploring marine resources and are rapidly becoming popular in both military and civilian fields.The key to ensuring the normal and stable operation of underwater vehicles is accurate underwater navigation and positioning.Due to the unique nature of the underwater environment,vehicles cannot obtain GPS signals,and a single navigation system cannot meet the requirements.Global researchers are studying integrated navigation technology and exploring new ways for underwater vehicle navigation and positioning.In order to improve its performance and accuracy,this article focuses on the application background of underwater unmanned aerial vehicles,adopts SINS/DVL integrated navigation system,and conducts digital simulation and pool experiments on the proposed navigation algorithm.The main research content includes:Introduced common coordinate systems in navigation,explained the working principle of strapdown inertial navigation,provided equations for solving attitude,velocity,and position,and established an error model on this basis;Starting from the Doppler effect,the working principle of DVL,including single beam,double beam,and four beam configurations,was introduced,and an error model was established.Based on the characteristics of Kalman filtering theory,the unscented Kalman filtering(UKF)in nonlinear discrete Kalman filtering was analyzed.The traditional UKF algorithm has relatively low filtering accuracy and is prone to divergence.A new adaptive filtering algorithm(RHAUKF)is proposed to address this issue.This algorithm constructs an estimation model for system noise based on the maximum likelihood criterion,and then introduces rolling time domain estimation for optimization.The Newton Raphon algorithm is used to solve the maximum likelihood estimation for noise statistics,ultimately obtaining an adaptive UKF algorithm.A SINS/DVL integrated navigation system was designed,using indirect filtering to provide its state equation and measurement equation,and a filtering model was established.Based on the established model,the proposed RHAUKF algorithm and the other two algorithms were used to simulate and analyze the integrated navigation.By comparing the attitude angle(heading,pitch,roll)error,three-dimensional(east,north,sky)velocity error,and threedimensional position error under the three algorithms.The results show that the designed SINS/DVL integrated navigation system can achieve high-precision navigation and positioning of underwater robots;The proposed RHAUKF algorithm can effectively suppress the interference caused by unknown or imprecise noise statistics.Finally,a pool experiment was conducted to verify the practicality of the integrated navigation system and the effectiveness of the proposed filtering algorithm.