The Research On Adaptive Path Planning In Complex And Heterogeneous Environment

The AUV, Autonomous Unmanned Vehicle, is a kind of intelligent underwater equipment which can autonomously sail at the sea without people,and the route planning is a key technology to realize its autonomous navigation. The AUV route planning belongs to the area of the three dimensional space, with high dimension and large amount of calculation and other technical difficulties. The existing methods are mainly by limiting the movement direction of the nodes which is to convert the three-dimension space into two-dimension space,and thus the difficulty of the route planning. But for complex heterogeneous environment,limiting movement direction of the nodes will bring about some problems,for example, the waste of storage space, and poor adaptability to the environment, so the study of the route planning method in the complex heterogeneous barrier environment has a very practical significance.According to analyzing and summing up the characteristics of the actual sailing route,we can find that the adjacent nodes of the route are not independent of each other, and there exists an interaction force between them. Based on the mathematical modeling of the interaction force and combined with the Particle Swarm Optimization (PSO) algorithm,this paper proposes a new route planning method. The main work accomplished in this paper is as follows:Firstly, this paper introduces the relative knowledge of the Particle Swarm Optimization algorithm, analyzes the space uniform-divided molding technology, and summarizes the deficiency existing in the environmental modeling method.Secondly, in view of the problems existing in the space uniform-divided molding technology and by observing the characteristics of the actual sailing route, this paper puts forward the elastic correlation property, and carries on its mathematical modeling, and gives the calculation formula of the elastic correlation operator. At the same time, this paper expounds the self-adaptive property and the function in reducing dimension about the elastic correlation property.Then, with adding the elastic correlation operator to the updating formula of the PSO,this paper puts forward a route planning method named the PSO based on the elastic correlation operator. In view of the two cases of the path intersects with obstacles, namely the nodes are in the obstacles or the adjacent nodes are not in the obstacles while their attachment intersects with the obstacles, this paper is to take corresponding obstacle-avoiding strategy respectively. According to the specific form of the fitness function and the control method of the node step, we are to finish the simulation experiment of the PSO based on the elastic correlation operator in different two-dimension obstacle environment. The convergence and convergence speed of the algorithm is analyzed. Compared with the PSO based on the space uniform-divided method, we can know the effect of the PSO based on the elastic correlation operator on the route planning in the complex heterogeneous environment.Finally, this paper applies the PSO based on the elastic correlation operator to the three-dimension space, and completes the route-finding experiment respectively in the spherical obstacle environment and the real underwater environment, and verifies its ability to find the route in three-dimension space. Because the real underwater topography surface is irregular, and is unable to be expressed with the mathematical formula, so this paper puts forward a method of inserting virtual nodes into the adjacent nodes to judge that the attachment of the adjacent nodes intersects with the obstacles whether or not.