Research On Collision Avoidance Algorithm Of Intelligent Ship Based On Risk Assessment Model Of Quaternion Ship Domains

With the development and utilization of marine resources and the prosperity of maritime trade,maritime collision has always been an important issue affecting the safety of ship navigation.Ship collision accidents not only cause huge loss of life and property,but also cause catastrophic damage to the marine ecological environment.Although the International Maritime Organization(IMO)has formulated the International Regulations for Preventing Collisions at Sea(COLREGs)to guide collision avoidance operations under different encounters,it is not easy for maritime personnel to comply fully with COLREGs rules through their own operations.In the complex multiple targets marine environment,it is difficult for the crew to make a correct judgment quickly.In order to solve this problem,this paper proposes a new calculation method of collision risk combined with ship domain and velocity obstacle algorithm,and solves the collision avoidance strategy suitable for the current situation based on genetic optimization algorithm for the complex encounter environment.The main work of this paper is as follows:(1)The relative knowledge of ship collision avoidance decision-making is summarized and the basic operation of ship collision avoidance is analyzed.Combined with the COLREGs rule,this paper analyzes the collision avoidance responsibilities of ships in various encounter situation,and determines the corresponding collision avoidance responsibilities of ships in complex environments with multiple ships.(2)A risk discrimination model based on ship domain and velocity obstacle method is proposed.Considering the ship's own parameters,the characteristics of the maneuvering level and the advantages of the velocity obstacle method in dealing with the nonlinear multi-objective problem,the collision risk of ships in complex environment is identified.At the same time,based on this model,a new method of calculating the risk degree is proposed,which uses the collision risk degree to quantify the collision risk degree between ships.(3)Based on multi-objective improved genetic algorithm,a step-by-step coordinated collision avoidance planning method is proposed.Considering the complexity of multi-objective encounter environment,a bi-criterion multi-objective genetic algorithm combining Pareto evolution and non Pareto evolution is used to solve the collision avoidance route.By improving the renewal mode of non Pareto evolutionary population,the convergence rate of the population is accelerated and the diversity of the population is increased.In view of the situation that other ships may take measures to avoid collision,the reasonable collision avoidance route of each ship is determined by using the communication equipment of the ship and step-by-step coordination method.According to the current route information and the planned collision avoidance route information,an improvement function is proposed to measure the advantages and disadvantages of the collision avoidance route.In this way,the collision avoidance trajectory can be determined,so that there is no danger of collision between ships.(4)In the MATLAB environment to achieve the establishment of the ship collision avoidance decision support platform,and in the setting of the scene to carry out simulation experiments and analysis.In view of three kinds of encounter situations between ships,the scenarios of two ships encounter,multiple ships encounter and emergency avoidance of ships under unexpected circumstances are simulated.The results of collision avoidance planning are presented through GUI graphical interface,and the distance between ships in the process of collision avoidance navigation is shown.The simulation results show that the proposed algorithm can solve the safe and economical collision avoidance path under various encounter scenarios.