Research On Unmanned Surface Vessel Path Planning Based On Improved Genetic Algorithm And Collision Avoidance Rules

The intelligent unmanned surface vessel（USV）can effectively reduce the probability of ship collision accidents caused by human factors,and plays an important role in both military and civilian fields.Intelligent decision-making is the core of autonomous navigation of the USV,of which path planning is the key technology,including route planning and dynamic collision avoidance.Route planning is to plan a collision-free path based on the known environmental information and the input information.Dynamic collision avoidance is to formulate collision avoidance strategies based on real-time detection information of dynamic obstacles.On this basis,this paper studies the path planning of the USV,that is,route planning and dynamic collision avoidance methods.The main contents are as follows:（1）The USV path following control algorithm based on integral backstepping controller and Line-Of-Sight guidance controller is researched.An underactuated threedegree-of-freedom USV model considering the environmental interference of wind,waves,and currents is established,and the heading controller and speed controller are designed based on the design of the Line-Of-Sight guidance algorithm,the nonlinear disturbance observer and the backstep control method based on integral action,so as to realize the path following control of the underactuated USV and provide a simulation environment for USV path planning.（2）The route planning method based on an improved genetic algorithm is studied.The restricted random method is designed to initialize the population to improve the quality of the initial population.The Pareto-binary tournament is used as the selection operator to select the population according to the smoothness,the path length and the security,so that the number of selected populations is always kept constant.The adaptive crossover and mutation probabilities are designed to overcome the "precocity" of the population;The delete operator is introduced to delete useless path points,and the security operator is introduced to keep the path within the security range.Taking the shortest path and the best smoothness as the fitness function,an USV route planning system is established.（3）The multi-ship collision risk assessment method based on dynamics ship domain model is studied.The dynamic ship domain model is established by considering factors such as ship maneuverability,ship length,ship width,the relative speed of the encountering ship and environment.The minimum encounter distance,the minimum encounter time and the relative distance are used as collision risk assessment indicators,thus a risk assessment model based on the dynamic ship domain is established to analyze the collision risk during the navigation of the USV.（4）The USV dynamic collision avoidance method based on the dynamic ship domain risk assessment model is researched.Based on the International Regulation for Preventing Collision at Sea（COLREGs）and the dynamic ship domain risk assessment model,the combination of the set-based guidance method and the velocity obstacle method is used for collision avoidance decision-making and path planning.The motion control,route planning and dynamic collision avoidance simulation model of USV are established by MATLAB/Simulink.Simulation results show that the designed path planning method can effectively improve the dynamic collision avoidance ability of the USV in real-time path planning on the basis of meeting COLREGs.