| With the rapid development of economy and society,various ocean-related activities such as ocean transportation,marine fisheries,and offshore drilling are becoming more frequent,marine vessels and personnel often encounter accidents in china.In the process of maritime search and rescue(SAR),how to quickly respond to an accident and formulate a scientific emergency response plan are the key factors that affect the efficiency and success rate of SAR.At present,the marine SAR plan is mainly relied on drift prediction model and SAR experience,which is subjective and inefficient.The proposed SAR planning methods also have shortcomings such as simple task assignment and difficulty in global optimization of SAR force scheduling.In order to solve the above-mentioned problems,this paper proposes a maritime SAR planning method based on optimal search theory,which optimizes maritime SAR decision-making from three aspects:search area determination,SAR force scheduling,and search task allocation.The main content and work of this article are as follows:(1)The optimal determination of the search area.In this paper,the search area is determined by the method of random particles.First,the drift trajectory of the SAR target is predicted by the drift prediction model.Due to certain errors in the prediction of environmental data such as wind,waves,tides,and currents,the Monte Carlo method is used to generate random Particles to quantify their uncertainty impact on target drift.Then,the optimal area containing the predicted particles is determined by the Graham algorithm and the minimum area bounding rectangle(MABR)generation algorithm.Finally,the area is divided into grids and the probability of containment are calculated to realize the prediction of the spatial probability distribution of the SAR targets over time.(2)Construction of maritime SAR force scheduling model and design of solving algorithm.This paper abstracts the maritime SAR force scheduling problem into a ship combination optimization problem,and builds a SAR force scheduling model based on the optimal search theory.The model aims to maximize the probability of successful search and rescue(POSSAR),fully consider the ship position,speed,search capability,rescue level and other factors,combined with ship number constraints and search time constraints,to optimize the SAR force scheduling plan.(3)Allocation of maritime search tasks.In order to overcome the problems of existing task allocation algorithms,including lack of task priority,overlap of task areas,and global coverage of tasks,this paper designs a regional task allocation algorithm that takes into account the characteristics of time and space.Based on the optimal search theory,the overall division rule is introduced,and the area is divided in stages according to the solution of the SAR force scheduling model,combined with the task area constraints and overlap constraints of each ship,and the optimization of SAR tasks in space and time is realized.(4)Algorithm verification and application.Taking the "BI HAI 159" round collision accident on June 27,2018 as an example,a SAR decision simulation experiment was carried out,and it was found that the SAR force scheduling solution based on the algorithm solved in this paper reduced the number of ships from 4 to 3 compared with genetic algorithm,and the POSSAR is significantly improved.Compared with the traditional method of assignment,the search area of the ship is expanded by 59%within the maximum survival time of the fallen person,which is 3.863h,and the first-coming ship can search the high probability area first.The above analysis proves that the proposed maritime SAR planning method not only saves costs,but also improves SAR efficiency and POSSAR,so it has obvious optimization.The maritime SAR support decision-making function module developed according to the method of this article was integrated into the"National Maritime Search and Rescue Support System(www.marinesar.cn)" in October 2019,providing decision support services for the national maritime SAR related departments. |
