摘要
以福建省公路旅客周转量和货物周转量的统计资料为基础,结合神经网络技术原理,应用BP神经网络方法建立3维输入、单输出、隐层单元数为15的3层神经网络模型,分别对福建省公路旅客周转量和货物周转量进行预测.结果表明,各月的旅客周转量和货物量预测值的最大相对误差的绝对值分别为0.4890%和0.4495%.该模型具有简便实用、预测精度高的优点.
Based on the data of highway passenger-kilometers and freight-kilometers in Fujian Province, the neural network model with three inputs and single output were established, and highway passenger-kilometers and freight-kilometers were predicted. The results showed that the highest predicting accuracy of the passenger-kilometer and freight-kilometer in each month were 0.4890% and 0.4495%, respectively. The model was easy, practical and feasible, which also had high accuracy.
出处
《福建农林大学学报（自然科学版）》
CSCD
北大核心
2007年第1期110-112,共3页
Journal of Fujian Agriculture and Forestry University:Natural Science Edition
基金
福建省自然科学基金资助项目(S0650004)
福建省教育厅资助项目(JA03075)
关键词
交通工程
BP神经网络
旅客周转量
货物周转量
traffic engineering
BP neural network
passenger-kilometer
freight-kilometer
