A Test And Evaluation Method For Ship Autonomous Collision Avoidance Performance In Complex Scenarios

Maritime transportation has always been the main mode of transportation in international trade.As the carrier of maritime transportation,the development of informatization and intelligence of ships has been paid more and more great attention globally.Intelligent ship is one of the hot spots in shipbuilding and shipping industry,and its development will bring great innovation and change to the maritime transportation mode.In order to make the ship operate intelligently until it finally replaces the traditional ship,its function and performance must be tested repeatedly.The ship autonomous collision avoidance system is the most core part of the intelligent ship navigation system,and its performance testing and evaluation methods are crucial to the development of intelligent ships.This thesis summarizes the ship collision avoidance test scenarios,proposes evaluation indexes for the performance of ship autonomous collision avoidance based on relevant information and expert suggestions,establishes the evaluation index system of ship autonomous collision avoidance performance in three aspects of safety,economy and efficiency,and determines the calculation methods of the corresponding index factors of the index system.In this thesis,the fuzzy comprehensive evaluation method is applied to construct the evaluation model of ship autonomous collision avoidance performance,and in the evaluation process,the indicators for which data can be directly obtained are evaluated by quantitative methods;for the indicators for which data cannot be obtained,the evaluation is carried out by a combination of qualitative and expert scoring methods,and the subjective weights are determined by using the sequential relationship analysis method,and the objective weights are determined by using the simple correlation function method.Then,the game theory method is introduced to consider the game relationship between each factor and get the final combination weights to reduce the deviation of subjective and objective weights,so that the subjective and objective weights can be scientifically and reasonably combined optimally,and then the fuzzy comprehensive evaluation model of ship autonomous collision avoidance performance is constructed.By collecting the simulation test data of the ship autonomous collision avoidance system developed by the author’s research team,the evaluation system and evaluation model of the ship autonomous collision avoidance performance constructed in this thesis are verified,and the performance index system and evaluation model constructed in this thesis are proved to be reliable and practical.This thesis realizes the quantitative evaluation of the ship autonomous collision avoidance performance,enriches the theory related to the testing of intelligent ship autonomous navigation system,and the results of this thesis have contributed to the development of intelligent ship testing technology.