Research On Path Planning Of Unmanned Surface Vehicle In Dense Obstacle Environment

Unmanned Surface Vehicle(USV)is a new type of ship,and owns the characteristics of autonomous navigation,so it generally becomes the research center around the world ship field.Path planning algorithm is the core technology of the USV,but now it meets a great problem that it is difficult to ensure the safety of USV navigation in dense obstacles.The issue strongly limits the scale application of USV.Therefore,this master thesis selects the path planning algorithm for USV in dense obstacles area.The ship collision risk model is a universal ship collision avoidance model.The model calculates the collision danger between two ships through the two ship’s parameters,and then formulates an avoidance strategy based on the calculation results and maritime rules,thereby ensuring the ship’s navigation safety.Based on this model,the thesis designs an USV-obstacles collision risk model.In this new model,two ship models are replaced by the USV and dense obstacles model,then we can get the USV avoidance strategy.Next the strategy is integrated into the specific path planning algorithm to implement engineering applications to improve the safety of USV sailing in dense obstacles.The specific content is as follows:Firstly,build the dense obstacles area model,the thesis changes the original single obstacle modeling mode.Specifically,the dense obstacle area is modeled as a whole to facilitate mathematical processing in subsequent algorithms.Such as the dense regular obstacles of the original regular circles or irregular polygons are densely remodeled into new regular circles or polygons,and next they are designed as new constraints of the algorithm.Secondly,calculate the collision risk of the USV and the dense obstacles area.The details are as follows: take the speed of the USV,the number of obstacles in the dense area,the average distance and other parameters as new influence factors,then add them to the original risk calculation formula,next grade the danger according to the calculation results,finally formulate corresponding obstacle avoidance Strategies: High-level danger immediately stop the ship,Medium-level danger to adjust speed and course,and Low-level danger to stay alert.Thirdly,take the obstacle avoidance strategies into different path planning algorithms for engineering improvement,including genetic algorithm,artificial potential field method,and A * algorithm.This step is to implement obstacle avoidance strategies Deep integration with path planning algorithms.Finally,simulation and experments were done to verify the validation of the improved algorithm,including GA and APF algorithm.Further,the thesis changes algorithm parameters and environment models to explore the general adaptation and optimal parameters of the algorithm,and the step provides a theoretical basis for further research.The experiment results show: the improved algorithms make the USV leaves far away from the high-level collision danger areas,and improve the navigation safety of USV in dense obstacles area.