基于GRNN的出租车保有量预测算法的应用

Application of urban taxi inventory forecasting algorithm based on GRNN

文章介绍了基于广义回归神经网络(Generalized Regression Neural Network,GRNN)的城市出租车保有量预测算法的特性和优点;并详细给出了其运用到某市的出租车保有量预测中的计算全过程,即出租车保有量的影响因素分析、网络的建立、训练和检测,以及最终模拟出的结果。通过应用表明,在特定条件下该方法可以有效提高预测的精度,为城市交通规划者提供合理的理论支持。

This paper introduces the characteristics and advantages of the urban taxi inventory forecasting algorithm based on the Generalized Regression Neural Network（GRNN）. The details of the entire process of its application to taxi inventory forecasting calculation of a certain city are described, including the analysis of factors affecting the taxi inventory, establishing, training and testing of networks, and the final results of the simulation. The practical application shows that under certain conditions the method can effectively improve the forecasting accuracy and provide reasonable, theoretical support for the urban traffic planners.