Marine traffic accident prediction based on particle swarm optimization-based RBF neural network

摘要

The future marine traffic accident situation is shown by using the marine traffic accident prediction method. Thus, marine traffic accident prediction method based on particle swarm optimization-based RBF neural network is presented in the paper. Particle swarm optimization algorithm, a kind of population-based optimization algorithm, is used to adjust the connection weights and the center and width of radial basis function. The marine traffic accidents of a certain terminal from 1996 to 2007 are applied to study the feasibility of the proposed PSO-RBF neural network. The comparison results between the proposed PSO-RBF neural network and normal RBF neural network can indicate that the prediction results of marine traffic accidents of the proposed PSO-RBF neural network are better than those of RBF neural network.