Multi-objective emergency resources allocation optimization for maritime search and rescue considering accident black-spots

Efficient and effective allocation of emergency resources is a premise for in-time maritime search and rescue (MSAR). In order to improve the allocation efficiency of maritime emergency resources and the performance of MSAR operations, an optimization model of emergency resources allocation is proposed, considering multiple restrictions including accident black-spots, possible locations of rescue bases, different types of emergency resources and rescue ships. In the first step, Long Short-Term Memory (LSTM) is applied to predicting the number of possible accidents in the investigated area. K-means algorithm is then introduced to identify the centers of accident black-spots. Finally, the multi-objective optimization is constructed and Elite-preserved Genetic Algorithm (EGA) is used to determine the optimum emergency resources allocation strategy. The results show that the allocation strategies can meet the requirements of the real-case with a budget utilization rate of 87.9%. The research can provide with useful reference for the risk management of maritime safety and MSAR practices. •A new maritime emergency resources allocation optimization model is proposed.•Long Short-Term Memory is used to predict the accident number.•Six black-spots are identified in case study by using K-means algorithm.•Different rescue ships for different types of maritime accidents are considered.