Research On AUV Multi-scale Integrated Navigation Algorithm Based On Wavelet Transform

Ocean has become the second largest strategic space of human survival and development after land.With the economic development and scientific research needs,Autonomous Underwater Vehicle(AUV)has been paid more attention as one of main equipment to explore and develop the ocean.AUV technology has made rapid development in the past 20 years.The high precision underwater navigation technology is one of the key technology and technical difficulties to be solved.Underwater navigation system provides AUV with real-time position information and attitude information,which is the basic guarantee for underwater mission and safe return.If the data with multi-scale features are analyzed on multiple scales,more valuable information can be obtained,and the effective information can be extracted at different scales and then merged,which can reduce the influence caused by uncertain factors.This paper from the actual requirement of high precision dead reckoning navigation system(DR)and BDS integrated navigation system,proposes a new method based on wavelet transform for AUV's integrated navigation system with multi-scale features and multi-scale features.Multi-scale Kalman filter algorithm effectively reduces the systematic error and improves the navigation precision.The main work includes:Firstly,this paper constructs AUV's DR/BDS integrated navigation system framework,analysis the main navigation sensor's working principle,and introduces dead reckoning navigation system theory and system error sources in details.Secondly,according to the underwater DR/BDS integrated navigation system's multi-scale characteristics,it proposes a multi-scale Kalman filtering method based on wavelet transform,which eliminates the influence of uncertain factors on the single scale and improves the filtering efficiency.Finally,the multi-scale Kalman filter based on wavelet transform and the classical Kalman filter are used to filter the DVL and FOG data in sea trials.The results show that the algorithm proposed in this paper has smaller navigation error and higher navigation accuracy.