基于AOSVR的交通流预测及参数选择

Traffic Flow Forecasting and Parameter Selection Based on AOSVR

实时、准确的交通流量预测是智能交通系统发展的关键.AOSVR是一种支持向量机的在线更新算法,具有模型在线学习的特点,可应用于交通流量的实时预测,其中模型参数的选择是预测性能的关键因素.利用大连SCOOT系统采集的实时数据,通过训练集求解AOSVR的不敏感损失系数ε和惩罚参数C,形成自适应参数选择的AOSVR方法.仿真结果表明该方法能够满足动态路网交通流量预测的实时性和精确性需求,具有一定的应用价值.

Real-time and accurate traffic flow forecasting is a key factor to the development of intelligent transportation systems（ITS）.Accurate Online Support Vector Regression（AOSVR）algorithm is introduced,which efficiently updates a trained SVR function whenever a sample is added to or removed from the training set,and it is believed that AOSVR will perform well for real-time traffic flow forecasting.However,the good generalization performance of AOSVR highly depends on good parameter selection（PS）.This paper describes simple yet practical approach to AOSVR parameter selection directly from the training data,using real traffic data of SCOOT system in Dalian city.Experimental and analytical results demonstrate the feasibility of applying AOSVR to traffic flow forecasting and prove that the AOSVR＇s parameter selection can better satisfy real-time demand of traffic flow forecasting and has good practicability.