摘要
利用一台国六柴油机进行热态WHTC循环试验,对循环中所有工况点,依据排气温度值和原始排气NOx体积浓度值,进行二值化状态标识(0或1)。以发动机转速、进气空气流量、燃油流量,功率和转矩作为输入变量,工况点二值化后的状态标识变量作为模型输出,构造逻辑回归分类模型。将循环所有工况点按照一定比例随机划分为训练工况数据点集与测试工况数据点集,训练逻辑回归分类模型。模型预测结果表明,测试工况数据点集对应的模型准确率和模型对标识为1的工况的召回率均大于83%。
Hot-started WHTC cycle tests were carried out on a ChinaⅥstage diesel engine.According to the exhaust temperature and NOx volume concentration of the exhaust gas without purification,all the operating conditions were marked with the binary state identification(0 or 1).A logistic regression classification model is built.The engine speed,intake air flow,fuel flow,power and torque are selected as the input variables,and the binary state identification of operating point is selected as the output of the model.According to a certain proportion,all operating conditions of the cycle are randomly divided into the training data sets and the testing data sets.The logistic regression classification model is trained.The prediction results of the classification model show that the accuracy of the model corresponding to the testing data sets and the recall rate of the model for the operating conditions identified as 1 are all larger than 83%.
作者
吉黎明
熊兴旺
杨子荣
JI Liming;XIONG Xingwang;YANG Zirong(China National Accreditation Service for Conformity Assessment,Beijing 100062,China;CATARC New Energy Vehicle Test Center(Tianjin)Co.,Ltd.)
出处
《小型内燃机与车辆技术》
CAS
2023年第2期6-9,20,共5页
Small Internal Combustion Engine and Vehicle Technique
关键词
工况
逻辑回归
阈值
准确率
召回率
Operating condition
Logistics regression
Threshold
Accuracy rate
Recall rate
