Application in Dynamic Path Selection for Emergency Vehicles Based on Hybrid Cuckoo Algorithm Optimizing Neural Network Model

Abstract

If any emergency occurs in a city, emergency vehicle scheduling must be such as to shorten the path travel time. The biggest difficulty in scheduling is how to measure the real-time changes in the traffic conditions of urban road network. To study the dynamic path selection of emergency vehicles under city emergency, this paper abstracts the urban road network into a map composed of nodes and edges, and takes the shortest path as the optimization objective. Firstly, the author takes the K-nearest sample set from the similar historical sample sets, predicts the real-time vehicle speed and establishes the path travel time function. Then, the author uses path reliability to measure the impacts of real-time traffic conditions on the overall travel time and constructs the two-stage objective optimization model for dynamic optimal path selection. Finally, based on this model, the author proposes a hybrid cuckoo search algorithm and uses it to optimize the weights and thresholds of neural network model to solve the K shortest-time paths in the dynamic road network, and take a partial road network in Yangpu District of Shanghai as an example for simulation test. The test results show that the proposed dynamic path selection model can reflect the actual scenario of emergency vehicle scheduling under emergency, and that the neural network model based on the hybrid cuckoo search algorithm is used to train weights and thresholds, so that the algorithm has a fast convergence speed and can solve the problem well. Compared with the classic cuckoo search algorithm and the particle swarm optimization algorithm, this algorithm has better performance.