Research On Real-Time Path Planning System For Unmanned Surface Vessel

With the increasing popularity of unmanned platform in the industrial,military and other fields,as an important tool for monitoring the marine environment,maintenance of maritime rights,Unmanned Surface Vessel(USV)has received great attention.Path planning systems is one of key technologies of unmanned autonomous control research.In this thesis,taking into account the actual navigation environment,through the combination of global and local path planning,the hybrid path planning method is studied.Based on the USV mathematical modeling,designing and implementing an experiment platform for online parameter identification of USV motion model parameters,the model parameters about "Lanxin" USV and the steering are obtained by identification.The global path is obtained by evolutionary genetic algorithm.According to the characteristics of the USV,the path is optimized by Bessel curve.Finally,the path planning problem of multiple static obstacles is solved.But there are also dynamic obstacle environments.And,Global path planning is the premise of a local path planning algorithm based on the velocity decomposition.Through real-time analysis and comparison the relative speed between USV and dynamic obstacles,the real-time Collision avoidance path can be obtained.Taking into account the USV motion state,combined with dynamic collision avoidance algorithm and the trajectory control,the USV real-time path planning system is feasible in theory.In the local simulation environment of single and multiple dynamic obstacles,the USV can avoid these obstacles and return to the global path,which verifies the feasibility of real-time path planning system by simulation.