应用神经网络优化高速铁路动车组周转的研究

Based on the domestic and overseas correlative research of high-speed train-set schedule, the paper introduces mathematic description and neural network model of the high-speed train-set scheduling. Then the thesis uses AINN to solve the super combinatorial optimization problem and analyses the theory course of using Hopfield neural network. After then the discourse brings forward sorting and clustering idea combining with simulated annealing to ameliorate times complexity. This method reduces the stay time and improves the efficiency of train-set schedule. In a word, the study provides the theory of optimizing and using computer to workout the train-set scheduling of the high-speed or special passenger railway. At last the article combines the actual special passenger line and use computer simulation to approve the feasibility and practicality of the study.