The timeliness and severity of maritime search and rescue (MSAR) determine that the resource selection should be fast and optimized as much as possible. At present, there is no intelligent optimization method to assist decision-making, which is suitable for the actual characteristics of MSAR. In this paper, the optimization problem of MSAR resource selection is abstracted into a multi-objective optimization problem. Considering the needs of actual MSAR resources, the response threshold model is introduced to improve the ant colony algorithm. It can effectively solve the situation that the ordinary optimization algorithm is easy to fall into the local optimal solution and ignore the better resources. The research work of this paper is divided into three parts: model construction, model solution and model verification. Firstly, a multi-objective optimization model with five practical constraints is constructed to minimize the search time and maximize the average utility of resources. Then, the response threshold model considering the actual resource demand is used as the heuristic information in the ant colony algorithm. The stimulus model and threshold model in the response threshold model represent the ability of resources to perform tasks and the threshold of resources to perform tasks respectively, so as to improve the efficiency and speed of the algorithm. Finally, the algorithm is verified and compared through an example. The experimental results show that the average resource utility of this algorithm is 20.6% higher than that of the basic ant colony algorithm, and the MSAR success rate is also improved, which verifies the effectiveness of this method.