公交客流预测的神经网络模型

Forecast of Passenger Volume Based on Neutral Network

通过分析神经网络的作用机理和公交年客流量的影响因素,以城市人口、居民收入、生产总值等9个因素作为输入神经元,输出神经元为每年的公交客流量,建立了公交客流预测的径向基神经网络模型(RBF)和BP神经网络模型.以合肥市公交量的调查数据为例,对网络进行学习与训练仿真实验,结果表明所建模型具有较高的预测精度,效果较好.

The mechanism of neural network and influential factors of passenger volume are analyzed in the paper. A neural network（RBF） and BP neural network model are developed in which the input nerve cells includes 9 nerve clells such as city population, income and production total value et al, the output nerve cells ineluds passenger volume. By analyzing the historical data of passenger volume of Hefei learn and train simulation experiments of the and effective.