高峰时段道路交叉路口二氧化碳浓度预测

Carbon Dioxide Concentration Forecast at Peak Crossing

对交通高峰时段道路交叉路口的二氧化碳浓度的预测,能够有效治理交通污染问题。对路口二氧化碳浓度的预测,需要提取交叉路口二氧化碳浓度特征,通过高斯核函数得到原始样本数据的特征值。传统方法提取路口二氧化碳浓度变化规律,计算机动车的多种尾气排放量,但忽略了对原始样本特征值的求取,导致预测精度偏低。提出基于改进支持向量机的交通高峰时段交叉路口二氧化碳浓度预测方法。该方法提取各个交通高峰时段的交叉路口二氧化碳浓度特征,通过高斯核函数得到原始样本数据的特征值,给出各个交通高峰时段的交叉路口二氧化碳浓度变化规律,由此预测交叉路口二氧化碳浓度。实验表明,所提方法预测精度较高,为改善交通高峰时段交叉路口的机动车排放量提供了有效依据。

ABSTRACT ： In order to predict the carbon dioxide concentration at crossroad, a prediction method for the carbon di- oxide concentration at crossroads during peak time based on improved support vector machines is proposed. The fea- ture of carbon dioxide concentration at each peak time is extracted. Through the Gaussian kernel function, the eigen- value of original sample data is obtained, and the change regulation of carbon dioxide concentration at crossroads dur- ing peak time is given. Thus, the carbon dioxide concentration at crossroads can be predicted. Simulation results show that the proposed method has high prediction accuracy, which provides an effective basis for improving the vehi- cle emission at crossroads during peak time.