重型车能耗与排放等级预测模型的构建

Construction of a prediction model for heavy-duty vehicle energy consumption and emission level

通过车载终端采集车辆行驶数据,构建重型汽车排放与能耗等级预测模型。对车辆进行短行程划分并计算特征参数,将短行程内的瞬时油耗同其对应的油耗限值的乘积分布分区间统计,将统计结果通过K-Means聚类得到初始等级,以区间分布为输入、初始等级为输出,构建能耗等级评价模型,并通过贝叶斯优化算法进行参数优化。类似地,进行排放等级评价模型构建,并将2个优化后的等级评价模型同另外3种算法的优化模型进行对比验证,结果表明,通过XGBoost_BO构建的模型性能更好。

The vehicle driving data was collected through the on-board terminal,and the prediction model of the emission and energy consumption level of heavy-duty vehicles was constructed.The vehicle journey was divided into short trips and the characteristic parameters were calculated.The instantaneous fuel consumption in the short trip and its corresponding fuel consumption limit product distribution were divided into interval statistics,and the initial grade of the statistical results was obtained through k-Means clustering.Taking the interval distribution as input,the initial level as output,the energy consumption level evaluation model was constructed,and the parameters were optimized by the Bayesian optimization algorithm.Similarly,the emission level evaluation model was constructed,and the two optimized level evaluation models were compared and verified with the optimization models of the other three algorithms.The results showed that the model constructed by XGBoost_BO had better performance.