Research On Maritime-Rules-compliant Collision Risk Incidence Assessment And Trajectory Planning Algorithm Of Unmanned Surface Vehicle

With the increasing demand of marine technology development,as an intelligent surface operation platform,unmanned surface vehicle(USV)has the advantages of flexibility,scalability and low cost.It plays an increasingly important role in military and civil applications,and the research in its related fields has been paid more and more attention.The autonomous navigation system is the key to realize its intelligent autonomous operation.It is required to realize the search and calculation of the optimal path in the complex water environment,and meet the requirements of safety,feasibility and economy.This paper focuses on three key functions of autonomous navigation system: global trajectory planning,collision risk incidence assessment and local trajectory planning.By reviewing the relevant research results and technical difficulties,the relevant improvement methods are put forward.The specific work of this thesis includes:(1)A global trajectory planning method based on an improved RRT algorithm is proposed.Firstly,the environmental obstacles’ region are expanded to ensure the safe distance between the planned trajectory and the obstacles.Then an improved method is proposed to overcome the shortcomings of the traditional RRT algorithm,such as slow convergence speed and poor output path quality.By introducing the goal oriented mechanism of random tree search,the problem of expansion blindness is solved;by combining the bidirectional search and variable step strategy,the search efficiency is improved;the length of planned path is significantly reduced by removing redundant nodes optimization;the turning point of path is optimized based on the motion primitives method,and the traceability of planned path is improved.Finally,the complete algorithm implementation process is summarized,and the simulation and experimental results show that the improved algorithm can plan the global trajectory to meet the performance requirements of USV.(2)A method of collision risk incidence assessment based on fuzzy theory is proposed.Firstly,the relevant provisions of COLREGs are summarized to provide rules for the formulation of collision avoidance strategies.Then,the collision risk assessment model of two ship and multi ship encounter scenario is constructed;in the two ship encounter scenario,the method of subjective AHP is used to calculate the weight of the assessment index;in the multi ship encounter scenario,the weight determination method is combined with the maximum deviation theory,and based on the principle of subjective combined with objective and pessimistic assessment.Finally,a complete implementation process is proposed,and the effectiveness of the evaluation model is verified.(3)A local trajectory planning method based on improved VFH and DWA algorithm is proposed.Firstly,the trigger conditions of local collision avoidance stage are specified,and the proposed collision risk incidence assessment and local trajectory planning are combined.Then,under the constraints of maritime rules,maneuverabilit y and safety,the improved form based on the traditional VFH and DWA algorithm is proposed,the collision risk incidence value of dynamic obstacles is regarded as the key parameter of histogram construction,a method of obstacle direction sector calculation based on prediction is proposed and the core cost function is designed and constructed;the combined algorithm form is proposed.Finally,the complete algorithm architecture of this thesis is proposed,and the relevant simulation verification is completed based on MATLAB platform,and the relevant experimental verification is completed based on the USV experimental platform.