Research On Target Searching And Cooperative Hunting For Autonomous Underwater Vehicles

This dissertation is on the background of an important foundation research project, remote navigation and intelligent control technology for XXX. In allusion to a given case in which some autonomous underwater vehicles (AUVs) are used to carry out the task of region searching and cooperative hunting for an invasive target, some cooperation oriented technologies for AUVs, such as system architecture, path planning, obstacle avoiding, autonomous searching and target hunting, are deeply studied in the dissertation.Firstly, according to the architecture of the individual architecture of AUV and the need of the mission case, a hybrid hierarchical architecture is presented based on the traditional distributed architecture, which includes the task planning layer and the behavior controlling layer and greatly enhances the cooperative ability of AUVs.Secondly, the path planning method based on the principle of ant colony optimization (ACO) is studied to optimize the operating path and the collision avoiding method based on the principle of artificial potential field (APF) is presented to solve the problem of AUVs colliding with the environmental obstacles and some moving object (including the other AUVs and the invasive target) while AUV is moving in the given area. Both of these two methods are combined as a new strategy and the detailed essential factors are proposed including the rule of state transition, the heuristic factor and the pheromone updating rule. In order to enhance the responding speed of AUV in complex environment, an improved rapid ACO path planning method is presented.Additionally, a regional random searching strategy is posted to meet the need of large-scale operating, which automatically divides the searching area according to the amount of AUVs and the scale of the working area. The sub-goal point is calculated and selected according to the position info of all AUV members among the cooperative team. Therefore, all the AUV members are dispersed averagely in the area. Obviously it has the advantages of easy calculating and real-time processing. As for the effective target hunting problem, an optimal rounding up strategy is brought forth based on the idea of "potential point". A polar coordinates system is set up, which regards the target position as the axis pole and accordingly the target moving direction as the polar axis. By ranking AUVs polar angles orderly in the polar coordinates system, the hunting tasks are assigned optimality for AUVs. When the "potential point" lands into the obstacle area, AUV will give up its hunting task to adapt to the environment. Both the dynamic "potential point" calculating method and the staged hunting strategy could reduce the sensitivity to the moving target during hunting process.Finally, the simulation software of AUVs is developed, which adopts the known six-freedom dynamic model and a PID controller to simulate the movement of AUVs and control their attitude motion. By designing different mission cases with different environment configuration, all absolved research results are verified and demonstrated accordingly. The simulating results show that AUVs can quickly find the invasive target and successfully hunting it in large-scale area and they could be adaptive with the environment and autonomous operating to a certain degree.