Research On UUV Formation Planning Methods For Underwater Crash Target Search

In recent years,marine wrecks have occurred frequently,which makes the question that how to use the existing search equipment to carry out more efficient underwater search work has become an urgent problem to be solved.Unmanned Underwater Vehicle(UUV),as a multiplier of ocean power,is regarded as an important tool to solve this problem.Especially with the introduction and development of the UUV formation concept,it provides a new idea for underwater search work.Therefore,this paper will focus on the task planning between UUV formations,so that it can quickly and efficiently complete the search of underwater crash targets.In solving the UUV formation mission planning problem for underwater target search,this paper first divides it into two stages: wide-area rough sweep and local fine search according to the search process.Based on this,the research on UUV formation wide-area target search algorithm and UUV formation submarine fine coverage search algorithm was carried out.The work done in this paper is as follows:Firstly,the article analyzes the UUV formation’s search process for underwater crash targets.A search plan for wide-area search for black box and fine search for wreckage was formulated.The wide-area search work is carried out using the vocal characteristics of the black box and thus the task area of the core search is determined.The article also establishes a mathematical model of the elements involved in the search work.Relevant elements include search area model、 UUV formation architecture and kinematics model、 passive sonar model for searching black box signals、forward-looking sonar obstacle avoidance model for collecting environmental information、side-scan sonar model for overlay imaging and communication model between formations.Secondly,the wide-area search problem in the search process is described and analyzed.A UUV formation wide-area search algorithm based on improved bio-heuristic algorithm is proposed.The algorithm adds a bio-inspired neural network algorithm based on the grid method.The neurons are associated with the grid unit one by one,and the sonar reliability function of the grid unit is assigned to the neuron and the search point of the next moment is determined according to the output activity value of the neuron.In addition,for the vocal characteristics of black box,the search idea of active and passive sonar simultaneous work is proposed.According to the characteristics of this kind of search method,an improved algorithm with more engineering practicability is proposed.The algorithm enables the UUV formation to shorten the search path、reduce the number of turns、reduce energy consumption and realize the fast black box search of the UUV formation in the grid map.According to the search results,the task area of the core search phase is determined.The simulation results show the effectiveness of the proposed algorithm.Then,aiming at the problem of area coverage in the fine search stage,the idea of partition coverage search suitable for distributed system is put forward.The area coverage problem is divided into partition problems and coverage issues for research.The search area is initially allocated according to the Voronoi partition principle,and the redistribution of the area is realized by the task allocation algorithm based on the balance principle.The algorithm achieves the consistency of the load of each UUV task area.After partitioning,the coverage path planning algorithm based on the improved raster reliability function is used to complete the coverage of the UUV formation in a small coverage area.Finally,the paper completes the GUI design based on Matlab,and gives the system construction of UUV formation search underwater crash target.The system encapsulates the wide-area search process and the fine search process involved in this topic.Through the interface button to realize the call of the relevant function in the library,the simulation process of UUV formation search is completed.Different simulation cases are designed for different search phases to verify the effectiveness of the research.