基于组合模型的短时交通流量预测算法  

Short-term Traffic Flow Prediction Algorithm Based on Combined Model

作　　者：芮兰兰[1] 李钦铭 

机构地区：[1]北京邮电大学网络与交换技术国家重点实验室,北京100876

出　　处：《电子与信息学报》2016年第5期1227-1233,共7页Journal of Electronics & Information Technology

基　　金：国家自然科学基金创新研究群体科学基金(61121061);国家自然科学基金(61302078;61372108);北京高等学校青年英才计划项目(YETP0476)~~

摘　　要：交通流量预测是实现智能交通技术的核心问题,及时准确地预测道路交通流量是实现动态交通管理的前提,短时交通流量的预测是交通流量预测的重要组成部分。该文针对十字路口的短时交通流量预测问题设计了基于交通流量序列分割和极限学习机(Extreme Learning Machine,ELM)组合模型的交通流量预测算法(Traffic Flow Prediction Based on Combined Model,TFPBCM)。该算法首先采用K-means对交通流量数据在时间上进行序列分割,然后采用ELM对各个序列进行建模和预测。仿真实验证明,与单一的BP(Back Propagation)神经网络和ELM相比,该组合模型算法建模时间为BP的1/10,ELM建模时间的4倍,均方误差为BP的1/50,ELM的1/20,该组合模型算法决定系数R2更接近于1,模型可信度更高。Traffic flow prediction is a key problem of realizing intelligent transportation technology. Forecasting traffic flow in time and accurately is the precondition to realize the dynamic traffic management. Short-term traffic flow prediction is an important part of traffic flow prediction. In this paper, the Traffic Flow Prediction Based on Combined Model（TFPBCM） based on traffic flow sequence partition and Extreme Learning Machine（ELM） is designed for the short time traffic flow forecasting. The algorithm divides the traffic flow into different patterns along a time dimension by K-means, and then models and forecasts for each pattern by ELM. The proposed algorithm is compared with Back Propagation（BP） and ELM. The combined model algorithm on modeling time is 1/10 of BP, but is 4 times ELM. Its MSE is 1/50 of BP and 1/20 of ELM. The combined model algorithm＇s coefficient of detemination（R2） is close to 1, so the credibility of the model is higher than others.

关 键 词：短时交通流量 K均值算法 极限学习机 组合模型算法 

分 类 号：TP393.07[自动化与计算机技术—计算机应用技术]