Emergency vehicle path planning for university campus traffic based on reinforcement learning cuckoo search algorithm

In this paper, a novel emergency vehicle path planning approach tailored for university campus traffic is introduced, leveraging reinforcement learning combined with the cuckoo search algorithm. Firstly, an evaluation index system for campus traffic conditions is established, employing the analytic hierarchy process and expert evaluations to assess the prevailing traffic scenarios. Based on these assessments, an objective function is formulated specifically for emergency vehicle path planning within university campuses. Subsequently, the reinforcement learning cuckoo search algorithm is applied to solve this objective function, yielding an optimal path planning strategy. Experimental results demonstrate the efficacy of the proposed method. It achieves a vehicle detour coefficient ranging between 0.01 and 0.13, with an average vehicle travel distance of 5.34 kilometers and an average path planning time of 1.38 seconds. These findings underscore the method's capacity to significantly improve path efficiency and reduce planning time for emergency vehicles navigating university campuses.