Research On Multi-AUV Target Search And Positioning In Unknown Environment

Nowadays,countries are increasingly fierce in their development and competition in the oceans.In the face of the unknown and complex marine environment,the search and positioning of underwater targets has been continuously studied,and the Autonomous Underwater Vehicle(AUV)is used as a new type of underwater equipment with high flexibility and strong autonomy is an important means to solve underwater problems.This article focuses on the search and positioning of multi-UUV cooperative target searching in an underwater environment that is completely unknown:Firstly,we analyze the key and difficult issues of target searching and positioning in an unknown environment,analyze the movement restrictions in detail of AUV by building an grid model of environment,kinematic model of AUV,forward-looking sonar model,and feature model of target.Effects of noise and visual thresholds in sonar.Secondly,to analyze the AUV target positioning problem,this paper uses Bayesian probability estimation to estimate the position and status of AUV.In multi-AUV system,after each target is observed several times,the initial observation is based on the prior information.The target observation sample information is continuously added to obtain the posterior probability estimation,so the positioning error can be continuously reduced and the positioning accuracy can be guaranteed.Before the estimation of the target position,it is also necessary to accurately know the positioning of the AUV and use the Kalman filter to reduce the influence of the AUV motion noise when designing the route for the AUV,select the UKF application with better filtering effect by comparing the UKF and the EKF.Then,taking into account the timeliness of the search method,This article propose a study based on perceptual adaptive target search method using sub-region search strategy.The adaptive search method is to search the track for AUV online design,compared to other off-line design optimization.In terms of trajectory,it can ensure positioning accuracy while avoiding missing key information.Because the optimization algorithm itself has a serious dependence on environmental data,according to the sonar model parameters,the large-scale sea area is divided into several sub-areas.In the case of no target within the sight-range,sub-area search strategy can be used to multi-AUV system.The cooperative relationship among them improves search efficiency and ensures effective implementation of target searching and positioning.Considering that there are threats of uncertain obstacles in the underwater environment,according to the principle of timeliness,the improved dynamic window method is applied to obstacle avoidance.Finally,the simulation verifies the feasibility and adaptive search capabilities of the above methods.Finally,the above target searching and locationing method is applied to a three-dimensional unknown underwater environment.According to the sonar parameters,the search layers are divided according to different depths,and each search layer utilizes a sub-area search strategy,and multi-AUV systems can perform different searches.The search state of the layer and the state of the search layer where it is located.After the optimization calculation,the search layer with different depths is asynchronously switched,so that it can cooperate to complete the target searching and positioning task.The simulation verifies the environmental adaptability and search efficiency of the multi-AUV system in the presence of faulty AUV and dynamic targets.