Research Methods Of Avoiding UUV's Moving Obstacles In Dangerous Environment

When unmanned underwater vehicle(UUV)navigates to carry out mission in the marine environment,all kinds of obstacles will interfere with UUV's navigation,not only include the static obstacles such as islands,reefs and other,also include the moving obstacles such as ships and other.Avoiding obstacles in security and arriving at the designated area successfully puts forward the higher request to the UUV.When the UUV navigates in unknown waters,making decision of the collision avoidance is inaccurate and incomplete,due to its own state and the environmental perception of uncertainty.Therefore,this article will research UUV's avoidance methods of avoiding obstacles in dangerous environment.First of all,through the analysis the characteristics of forward looking sonar data,a lot of abnormal data include in forward-looking data,due to various interferential factors such as noise.The reliability of forward-looking sonar data acquisitied in the complex underwater environment is poor,and can not meet the safe demand when UUV navigates to avoid static and dynamic obstacles.So using the wavelet transform deal with the data obtained from forward looking sonar eliminates noise.Using the pre processed data completes the construction of obstacles outline by support vector clustering,so as to obtain the approximate location of obstacles,to lay the foundation for the subsequent avoidance.Secondly,the position and state estimation of the moving obstacles is the important implementation of UUV to avoid moving obstacles.To solve the moving obstacles' s position uncertainty problems,we analysis of dynamic obstacles motion model,estimate and predicte the moving obstacles motion condition.Then we design variable dimension kalman filter based on kalman filter.We use variable dimensional kalman filter to estimate position motion state of moving obstacles.It can provide powerful help for real-time decision making of UUV.Again,analysis UUV behavior model of avoiding static obstacles and moving obstacles.So we can design the corresponding collision avoidance strategy.We use fuzzy control to avoid dynamic obstacles considering the risk of collision,so as to ensure the safety of UUV.Finally,considering the static and moving obstacles coexist situation,research UUV obstacles avoidance method By analysing of the advantages and disadvantages of artificial potential field method,we improve on it.By using UUV's resultant force in the improved artificial potential field,finish the avoidance of the static and dynamic obstacles.When using fuzzy control to avoid obstacles,the calculation is greatly increasing with the increase of the number of obstacles.the improved artificial potential field solves the problems.By using the improved artificial potential field,design real-time method to avoid obstacles to ensure UUV's in navigating.