基于联网收费数据预测行程时间的实现

高速公路海量的联网收费记录汇聚了很多极具价值的交通信息,利用数据挖掘方法,实现收费数据的增值利用具有重要的理论与应用意义。行程时间是交通规划与管理中的一项重要基础数据,以往行程时间的预测方法依赖于交通调查与检测数据,数据采集成本很高。本文以决策树理论建立起利用联网收费数据预测行程时间的思路,并借助SQL2005提供的数据挖掘功能,建立起数据分析系统,实现利用联网收费数据进行OD行程时间的预测。

Realization of an OD Traveling Time Forecast System Based on Highway Network Toll Data

Since the magnitude recorders of highway network toll data contain valuable traffic information, it is meaningful to develop the method to explore the use of toll recorder. Traveling time is the most importation basic data for traffic planning and management. Howev- er, the traditional method for traveling time forecast require the data form traffic survey or traffic surveillances, which result the high cost of data collecting. In this paper, the idea of forecasting the traveling time based on network toll data using Decision Tree theory is pro- posed, further the realization of the data analysis system in virtue of the utility of Knowledge Discovery in Database integrating in SQL2005 is briefly described. The system could achieve the traveling time forecast value using the highway network toll data.