Research And Application Of Coastal Ship Path Planning Method

Ships have always been the main means of transportation for global economic development due to its large carrying capacity,low freight costs,strong traffic capacity,and environmental protection.And ships have played a pivotal role in economic development.With the rapid development of maritime traffic,coastal waters have become one of the areas with the highest density of maritime traffic,which puts forward higher requirements for the intelligentization of maritime traffic.Path planning provides a general solution for ship intelligence.The existing coastal ship path planning methods cannot be applied to the actual marine environment due to problems that the planned paths are close to obstacles and have many inflection points.After an in-depth study of the current status of coastal ship path planning,firstly,this thesis proposes a global path planning method for coastal ships based on an improved Deep Deterministic Policy Gradient algorithm and an improved Douglas–Peucker algorithm.This method combines Long Short-Term Memory with Deep Deterministic Policy Gradient,and uses historical state information to approximate the current environmental state information,thereby making the predicted actions more accurate.At the same time,aiming at the problems of low learning efficiency and slow convergence speed of the model using the traditional reward function,this thesis designs main and auxiliary reward functions to optimize the reward principle of the traditional Deep Deterministic Policy Gradient.This not only helps to plan shorter and safer paths but also shortens the convergence speed of the model.Then,aiming at the problem that there may be too many turning points in the planned path,an improved Douglas-Pucker algorithm is proposed to optimize the path,making the final path safer and more economical.Secondly,this thesis proposes an improved hybrid A-star algorithm for local path planning.The exclusion gain is introduced into traditional hybried A-star to solve the problem that the planned path is close to the obstacles and ensure the safety of the path;meanwhile,the attraction gain is introduced to accelerate the speed and ensure the timeliness of path planning.Thirdly,in order to verify the effectiveness of the methods proposed,experiments were carried out in multiple simulation environments of varying complexity.The results show that the paths planned by the methods proposed in this thesis perform better in terms of planning time,navigation path security and economic cost than most other methods.Finally,combining with the electronic chart,the methods proposed in this thesis were applied in certain actual environments.It shows that paths planned by the methods proposed meet the navigation requirements of ships in the actual environment.