Research On Path Decision Making Method For AUV Based On ACO

This dissertation is on the background of an important foundation research item, remote navigation and intelligent control technology for XXX. It deeply studies the global path decision making method and the on-line path decision making method.In order to solve the problem of global path decision making, Firstly, the dissertation establishes an intervisibility graph model based on the traditional grid model, designs an status transfer updating rule and pheromone updating rule for ACO algorithm, proposes the operating steps and the detailed path decision making algorithm based on ACO which takes optimization of path as the object. Secondly, in order to make sure the path smooth, the dissertation designs a bidirectional cutting operator and an inserting operator to optimize the paths. Finally, according to the decision making problem which takes the length of path and the threat degree of path as the objects, a threat evaluating way, the threat-net concept and the process for setting up a share info table of threat source are proposed and introduced detailed in the dissertation.In order to solve the problem of obstacle avoiding, an on-line path decision making algorithm based on ACO is designed and realized for autonomous underwater vehicle. Firstly, it establishes the simulating model of multi-beam forward looking sonar based on gird model to detect the information of the unknown obstacles in the environment. Secondly, to overcome the disadvantage that in traditional grid model an ant can only select the adjacent grid as the next path point at the last step, it imports the concept of virtual intervisibility, so that it greatly advances.the efficiency of the ant colony optimizing process. Thirdly, to avoid AUV sharp turning led by on-line path decision making, a non-linear punishing mechanism is imported into the object function of ACO algorithm. In the end, the operating steps of AGO are proposed for on-line path decision-making for AUV.All above researched algorithms are tested by a simulating platform based on Visual C++. Lots of simulating results demonstrates that the path decision making method based on ACO is efficient and suitable for AUV, and it also has some outstanding advantages such as short searching time, optimized and safe path, and so on.