Font Size: a A A

Research On Collision Risk Identification And Navigation Strategy Of Ships In Open Water

Posted on:2021-05-04Degree:MasterType:Thesis
Country:ChinaCandidate:Z Y ZhangFull Text:PDF
GTID:2381330602989150Subject:Traffic Information Engineering & Control
Abstract/Summary:PDF Full Text Request
In recent years,intelligent ship has become a new hotspot in the international maritime industry,the Rules for Intelligent Ships and Action Plan for the Development of Intelligent Ships regard intelligent navigation system with autonomous navigation collision avoidance technology as one of the key technologies that need to be broken through in the development of intelligent ships.Collision is also the main type of ship accidents leading to casualties,and accounts for the main proportion of all maritime accidents.Based on the above background,this paper takes the open water as the research project,carries on the research method of ship collision risk identification and navigation strategy analysis,and intends to provide a new idea for the research and development of ship intelligent navigation system and ship collision avoidance.The specific research are as follows:(1)Aiming at the defects of traditional collision risk identification method that the target ship parameters are not considered comprehensively and the same safe encounter distance is used in different encounter situations,a ship collision risk identification method based on the ship domain model is constructed.The Kijima model is selected to define the range of ship safety encounters,the new decision parameters for ship collision risk identification is proposed:upper and lower boundaries of relative motion navigation,and the solutions of the decision parameters are derived.Combining with the geometry of KIJIMA model to divide the area of ship safety encounter,by comparing the numerical relationship between the relative motion course of the target ship and the boundary relative motion course in each area,the collision risk between ships could be effectively identified.(2)The ship navigation decision-making is regarded as a multi-objective optimization problem,and a fitness function model is constructed that takes into account the safety of ship navigation,the compliance of International Regulations for Preventing Collisions at Sea and the economy of avoidance process.The adaptive genetic algorithm is selected as the basic algorithm to solve the ship navigation strategy.In order to solve the problem of unstable convergence and easy to fall into the local optimal value of the adaptive genetic algorithm,an improved strategy of introducing the optimal preservation strategy and the Metropolis criterion into the traditional algorithm is proposed,and the Shubert function is selected as the test function to verify the effectiveness of the improved strategy.(3)By using the collision risk identification method based on the ship domain model and the ship navigation decision model based on the improved adaptive genetic algorithm,the collision risk identification and navigation decision-making processes of single ship encounter and multi ship encounter are constructed respectively,and the ship encounter case scenarios are set up for experimental verification.By comparing the risk identification results,the trajectory of own ship and target ships,the numerical relationships of ship distances and safety encounter distances,and the loss of voyage caused by the avoidance process verify that the method constructed in this paper could accurately identify the collision risk between ships,and the proposed navigation strategy could achieve the safe avoidance of the target ship.Facing with the needs of intelligent ship and maritime traffic safety,the collision risk identification method and navigation strategy analysis process of ships in open water studied in this paper could play an auxiliary supporting role in ensuring the safe and efficient navigation of marine ships.
Keywords/Search Tags:Ship Collision Avoidance, Collision Risk, Navigation Strategy, Ship Domain, Adaptive Genetic Algorithm
PDF Full Text Request
Related items