基于VMD-SSA-LSSVM模型的汽油车CO_(2)排放预测  

Prediction of CO_(2)emissions from gasoline vehicles based on the VMD-SSA-LSSVM model

作　　者：吐尔逊·买买提 成思怡 刘亚楼 TursunMamat;CHENG Siyi;LIU Yalou(School of Transportation and Logistics Engineering,Xinjiang Agricultural University,Urumqi 830052,China;Intelligent Transportation Engineering Research Center,Xinjiang Agricultural University,Urumqi 830052,China)

机构地区：[1]新疆农业大学交通与物流工程学院,乌鲁木齐830052 [2]新疆农业大学智能交通工程研究中心,乌鲁木齐830052

出　　处：《交通科技与经济》2025年第1期43-49,共7页Technology & Economy in Areas of Communications

基　　金：国家自然科学基金项目(51768071);新疆农业大学国家级大学生创新项目(202310758020);新疆农业大学交通运输工程校级重点学科开放课题项目(XJAUTE2022K01)。

摘　　要：以排放标准为国Ⅴ的轻型汽油车为研究对象,运用便携式车载尾气排放测量系统(PEMS)对实际行驶中污染物排放试验数据进行分析。先采用变分模态分解方法(VMD)进行分解降噪处理,再通过麻雀搜索算法(SSA)、最小二乘支持向量机(LSSVM)和长短期记忆网络(LSTM)的方法,构建一种高精度的轻型汽油车CO_(2)排放预测模型。通过随机森林和皮尔逊相关性分析,筛选出影响轻型汽油车CO_(2)排放的关键特征参数,构建数据集。利用VMD算法对数据集进行分解降噪处理,再采用SSA算法优化LSSVM模型,最终建立基于VMD-SSA-LSSVM的轻型汽油车CO_(2)排放量预测模型。结果表明:该模型对CO_(2)的预测精度和拟合效果均优于单一的LSSVM、LSTM以及VMD-LSSVM、VMD-LSTM、VMD-SSA-LSTM模型,能够为轻型汽油车CO_(2)排放预测提供参照;VMD-SSA-LSSVM模型在轻型汽油车CO_(2)排放预测方面具有显著优势和应用前景。The portable vehicle exhaust emission measurement system(PEMS)was used to analyze the emission test data of pollutants in the light gasoline vehicle with emission standard as Country V.Firstly,variational mode decomposition(VMD)was used for decomposition and noise reduction,and then Sparrow search algorithm(SSA),least square support vector machine(LSSVM)and long short-term memory network(LSTM)were used to construct a high-precision CO_(2)emission prediction model for light gasoline vehicles.Through random forest and Pearson correlation analysis,the key characteristic parameters affecting CO_(2)emissions of light gasoline vehicles were screened and the data set was constructed.The VMD algorithm was used to decompose and reduce the noise of the data set,and then the LSSVM model was optimized by SSA algorithm.Finally,the CO_(2)emission prediction model of light gasoline vehicles based on VMD-SSA-LSSVM was established.The results show that the prediction accuracy and fitting effect of the model are better than that of the single LSSVM,LSTM,VMD-LSSVM,VMD-LSTM,VMD-SSA-LSTM model,which can provide a reference for the CO_(2)emission prediction of light gasoline vehicles.VMD-SSA-LSSVM model has significant advantages and application prospects in CO_(2)emission prediction of light gasoline vehicles.

关 键 词：轻型汽油车排放 PEMS 随机森林 皮尔逊系数 VMD-SSA-LSSVM模型 CO_(2)排放预测 特征参数筛选 

分 类 号：U491[交通运输工程—交通运输规划与管理]