Research On Autonomous Planning Method Of Multiple Unmanned Underwater Vehicles Cooperative Executing Multiple Task

Multiple unmanned underwater vehicle (UUV) cooperative executing multi-task autonomous planning mainly includes two aspects of Multiple UUV cooperative executing multi-task allocation and UUV executing multi-task path planning. Multiple UUV cooperative executing multi-task allocation is that one or more large and complex tasks according to the task properties and certain rules decomposed into diifferent types of sub tasks,and then different types of sub tasks according to a certain order assigned to UUV with corresponding attribute.UUV executing multi-task path planning, which is after the completion of task allocation, according to the sequence of tasks assigned to each UUV,planning the navigation path for each UUV performing tasks in unknown environments. The main research work of this paper is as follows:Firstly,describe the types of task and UUV involved in UUV group cooperative multi-task autonomous planning and establish a model of autonomous planning. In the model,the relationship between UUV set, task set and task subset is described, and the evaluation function and constraint conditions are designed.Secondly, according to the different types of tasks and UUV, the task allocation is divided into two types of homogeneous UUV single type multi-task allocation and heterogeneous UUV multi-type multi-task allocation. Aiming at the homogeneous UUV single type multi-task allocation, this paper uses the improved ant colony optimization algorithm, designing the state transfer probability, task allocation evaluation function and algorithm flow. And This paper according to the principle of task attributes and UUV function matching,proposes transforming the heterogeneous UUV multi-type multi-task allocation to multiple homogeneous UUV single type multi-task allocation.Thirdly,Then, aiming at UUV in unknown environment according to the results of task allocation to perform the task,environment model is established based on multi beam forward looking sonar sight range. During the UUV navigation process, if the forward looking sonar detects obstacles in the navigation path, the specific location information of obstacles can be got by the model and then call the improved particle swarm optimization algorithm to generate a new security path, successfully avoid obstacles, thus realize the path planning in unknown environment.Finally, based on the Qt development environment, the development of simulation software for UUV group cooperative multi-task autonomous planning is completed. In this software platform, a comprehensive experiment is conducted on two aspects of multiple UUV cooperative executing multi-task allocation and UUV executing multi-task path planning. The experimental results show that the improved ant colony optimization algorithm and the improved particle swarm optimization algorithm are feasible and effective in achieving the task allocation and path planning in unknown environment.