Font Size: a A A

Research On Intelligent Ship Collision Avoidance Model Based On MCTS

Posted on:2020-08-19Degree:MasterType:Thesis
Country:ChinaCandidate:Q H ZhuFull Text:PDF
GTID:2392330602458494Subject:Nautical science and technology
Abstract/Summary:PDF Full Text Request
The occurrence of ship collision accident will not only cause great economic losses,but also cause serious damage to the ecological environment of the ocean and other water systems,and more importantly,it is easy to cause serious casualties.Therefore,reducing and preventing ship collision accidents has always been the most important issue in the maritime community,and the key problem is to reduce the judgment and operation errors of ship operators when avoiding collision,so as to reduce the occurrence of ship collision accidents to ensure the personal safety of crew and the safety of ship navigation.This paper is mainly based on Monte Carlo Tree Search algorithm,combined with different collision risk calculation methods,to establish intelligent ship avoidance model,to guide ship operators to make safe and effective ship avoidance decisions,and to ensure the safety of crew and ship.The main contents of this paper are as follows:Firstly,on the basis of the basic principle of ship collision avoidance,the method of calculating collision risk degree with single ship encounter is established by using the method of fuzzy mathematics,and determining the best priority of ship collision with multi-ship encounter is established by using the theory of variable fuzzy optimization combined with comprehensive weight.The experimental results show the effectiveness of the two methods.Secondly,the Monte Carlo Tree Search algorithm in artificial intelligence is used as the core algorithm of ship intelligent collision avoidance model.Based on the analysis of MCTS algorithm and its implementation algorithm(UCT algorithm),the applicability of MCTS algorithm is verified by comparing it with the classical greedy algorithm of decision-making optimization.Finally,considering the safety,economic factors of ship steering avoidance and the differences among different factors,a more reasonable return function which is suitable for MCTS algorithm is established.Monte Carlo Tree Search algorithm is used for decision-making optimization,and the optimal steering avoidance range is obtained.The intelligent collision avoidance of ships is simulated by using MATLAB software and the effectiveness of steering collision avoidance is verified.The final result shows that the algorithm has good applicability and can provide certain decision support for ship operators.
Keywords/Search Tags:Intelligent Ship Avoidance, Fuzzy Optimization Model, Collision Risk Index, Monte Carlo Tree Search Algorithms
PDF Full Text Request
Related items