Research On Path Planning Of Surface Unmanned Vehicle Based On Improved Whale Optimization Algorithm

Unmanned Surface Vehicles(USVs)are autonomous waterborne vehicles capable of independent operations,requiring the ability to interact with the external environment.To achieve this goal,USVs need to possess path planning and dynamic obstacle avoidance capabilities to address various potential hazardous situations.During the navigation of USVs,not only is global path planning decision-making necessary,but timely responses to local hazardous environments are also required to prevent accidents.Only by doing so can the USVs ensure safe,efficient,and successful mission completion.Therefore,the reliability of path planning algorithms and the ability to avoid dangers are crucial for the intelligent development of USVs,as well as the primary focus of this research paper.(1)Analyzing the current research status of path planning for unmanned surface vehicles(USVs)both domestically and internationally,it can be concluded that existing path planning algorithms have their own advantages and disadvantages.Currently,most navigation strategies focus on complex environments.However,the existing planning algorithms are inadequate to meet practical needs.Therefore,it becomes necessary to propose improved trajectory path planning schemes and obstacle avoidance strategies.Accordingly,this study is established with USV path planning as the fundamental design principle.(2)Establishing a motion model for unmanned surface vehicles(USVs)on the water.Considering the influences of wind,waves,and currents on USV path planning,a MMG(Mass,Moment,and Geometry)separation approach is adopted as a reference model.The USV is divided into longitudinal,lateral,and vertical forces to account for the effects of wind,water currents,and fluid dynamics acting on the USV.Mathematical models are developed to quantify the impact of wind on the USV,the influence of water currents on the USV,and the fluid calculations applied to the USV.(3)Design of a pairwise cruising waypoint path planning algorithm and optimization of waypoint ordering.Building upon the motion model for unmanned surface vehicles(USVs),this study begins by providing a brief overview of Whale Optimization Algorithm(WOA)and Ant Colony Optimization(ACO)algorithms,summarizing their respective limitations,and proposing improvement strategies.These strategies include initializing non-uniform distribution of pheromones,integrating transfer probabilities with whale updates,and enhancing pheromone update strategies to further enhance global search capabilities.Through simulation experiments comparing ACO with the improved algorithm,it is demonstrated that the improved algorithm achieves better and faster results.Regarding the optimization of waypoint ordering in the cruising path,an improvement is made to the update mechanism in the strategy-solving process of the WOA,allowing the algorithm to optimize non-linear strategies and provide optimal solutions.Through simulation experiments,the algorithm is shown to provide optimal solutions for nine cruising strategies,demonstrating its effectiveness.Finally,in real-ship experiments,excellent cruising trajectories are obtained,confirming the effectiveness and reliability of the algorithm.(4)Multi-objective path planning for unmanned surface vehicles.Building upon the waypoint-based path planning,this study proposes optimizing four objectives: path length,path smoothness,time cost,and safety.Mathematical models are established for these objectives,and corresponding objective functions and constraint functions are provided.In the context of multiobjective algorithms suitable for USV path planning,a DAMOACOWOA(Dynamic Augmented Multi-Objective Ant Colony Optimization-Whale Optimization Algorithm)algorithm is proposed.The DAMOACOWOA algorithm consists of the ACO-WOA module for extracting potential solutions,the intelligent search method module,and the archive update module for retaining,eliminating,and constraining the optimal solution solutions and their number.The related discernibility and path selection strategies are used to obtain the corresponding weightings between the objectives.Ultimately,the algorithm obtains the optimal path solution.(5)Research on dynamic collision avoidance decision-making algorithms for unmanned surface vehicles.Building upon global path planning,this study focuses on the research of dynamic obstacle avoidance decision-making in local paths.The problems that need to be addressed in collision avoidance are analyzed,and corresponding collision avoidance strategies are proposed.Based on deep reinforcement learning,the action space and state space are designed.Drawing upon the principles of Velocity Obstacle(VO),a deep reinforcement learning approach is proposed to solve the dynamic obstacle avoidance problem.By designing behavioral variables,state variables,and reward functions,the algorithm is trained using simulation testing and ultimately achieves good obstacle avoidance results in real-ship experiments.