Research On Integrated Localization Algorithm Of Autonomous Underwater Vehicle Based On Single Beacon,inertial And Vision

In recent years,autonomous underwater vehicles(AUVs)have been widely used in various complex underwater tasks,including seabed topography mapping,marine mineral resource surveys,and marine environmental monitoring.The accurate localization of the AUV is essential to ensure its own safety and the accuracy of the collected data.As the global positioning system signal cannot be used in the deep sea,AUV navigation and localization has become a more challenging problem compared to the localization of ground robots.Compared with traditional underwater acoustic localization systems,acoustic localization based on single beacon ranging requires only one beacon to be deployed,which saves significant time and cost for AUV localization systems,and is currently a research hotspot in AUV localization.This thesis will conduct in-depth research on the problems of AUV integrated localization based on single beacon ranging and propose solutions:A single-beacon-assisted inertial localization algorithm based on the fusion of extended Kalman filter and nonlinear optimization is proposed.To solve the problems of the two commonly used single-beacon-inertial combined localization algorithms:the classical extended Kalman filter(EKF)method uses only one range measurement information,and the virtual long baseline(VLBL)method ignores the measurement errors introduced by inertial navigation,the scheme is proposed.The proposed localization algorithm not only combines the advantages of simple and real-time performance of the EKF method,but also uses nonlinear optimization to process multiple range measurements,and by introducing a pre-integrated inertial measurement unit(IMU)measurement model,it considers the uncertainty of relative motion between the acoustic range measurements measured by IMU.Non-linear optimization provides the optimal relative position of a single beacon for EKF global state update.The simulation results show that the localization accuracy of this algorithm is improved to a certain extent compared with the traditional two localization algorithms,while meeting the requirements of reliability and real-time.A localization algorithm based on the fusion of single beacon ranging and visual odometry is proposed.In the VLBL method,the sensor that measures relative motion is mainly an expensive Doppler velocity log(DVL)or IMU.In some scenes such as near the seabed,the camera can also be used as a motion sensor.In the proposed localization algorithm,considering the characteristics of different update frequencies of acoustic and visual information,visual odometry(VO)is used as a bridge between adjacent acoustic range data,and automatically moves the reference point of VO to each new acoustic location.VO uses global optimization to find the optimal relative displacement,which is used in the solution of VLBL.The simulation experiment conducted through real underwater image data and simulated acoustic range measurement data verifies the feasibility of the proposed acoustic-visual combined localization algorithm.