The Research Of Global Path Planning Of Autonomous Underwater Vehicle

The work of this thesis stems from the project of "Research on Autonomous Underwater Vehicle (AUV)". In this thesis the motion planning system for AUV was established to deal with problems (such as the nonlinearity of the vehicle itself as well as the uncertainty of the environment) that AUV faces while carrying out tasks. The simulation results show that our motion planning system enables the AUV to work more intelligently.As a main component of autonomous navigation system, global path planning (GPP) ability was an important criterion for the intelligent level of the autonomous robots. The A Star algorithm was more efficient than the other algorithms in optimization of the overall path length as well as the maturity of the algorithm. In this thesis the improved A Star algorithm with changing step length was put forward.In this thesis development of the AUV as well as methods tackling with the problem of global path planning (such as genetic algorithm , artificial potential field, transmission of activated value) were introduced. Procedure of environment modeling using octree and method for neighbor area searching were introduced in detail in the thesis. And then, the theory, concept, merit and flow chart of the A Star algorithm were given in the thesis and improved A Star algorithm with changing step length was put forward. Impacts of the ocean current on the motion ability of the underwater vehicle were considered and the problem of yaw amplitude keeping was discussed. Efficiency and feasibility were verified through simulation experiment.The unit decomposing method was designed to tackle the problem of path planning in certain area in which the AUV aims to search for objects such as mines. Geometry method based on the information of sensors was put forward for detecting underwater objects. Application of the unit decomposing method the result of the path planning of area in concern can cover the whole area, In addition the environmental information of area under search can be exactly depicted as well.