基于短行程特征聚类的移动源排放清单构建方法-学报杂志社

期刊信息

合肥工业大学（自然科学版），2023年12月，第46卷第12期：1585-1594

DOI: 10.3969/j.issn.1003-5060.2023.12.001

作者信息

李兵兵 $ ^{1} $，康宇 $ ^{2,3,4} $，曹洋 $ ^{2,3,4} $，李亚民 $ ^{4} $，许镇义 $ ^{2} $

(1. 安徽省生态环境监测中心，安徽合肥 230071；2. 合肥综合性国家科学中心人工智能研究院，安徽合肥 230088；3. 中国科学技术大学自动化系，安徽合肥 230026；4. 中国科学技术大学先进技术研究院，安徽合肥 230088)

摘要和关键词

摘要: 文章基于主成分分析（principal component analysis，PCA）和 K-means 聚类的短行程法，构建福州市轻型车的行驶工况。首先对原始数据进行预处理、划分运动学片段并计算其特征参数，运用 PCA 对划分后片段进行参数降维，再对其进行 K-means 聚类分析，形成典型片段库；然后分别根据最小参数偏差法和最大相关系数法挑选片段组成行驶工况，对比后输出最终工况，并对最终行驶工况进行有效性验证；最后利用 VT-Micro（Virginia Tech microscopic）模型并结合合成工况，计算单车排放因子，构建 2020 年福州市机动车排放清单。研究结果可为移动源排放清单构建研究提供参考。

关键词: 行驶工况重构；主成分分析（PCA）；K-means聚类；最小参数偏差法；最大相关系数法

Authors

LI Bingbing $ ^{1} $, KANG Yu $ ^{2,3,4} $, CAO Yang $ ^{2,3,4} $, LI Yamin $ ^{4} $, XU Zhenyi $ ^{2} $

(1. Anhui Eco-environment Monitoring Center, Hefei 230071, China; 2. Institute of Artificial Intelligence, Hefei Comprehensive National Science Center, Hefei 230088, China; 3. Department of Automation, University of Science and Technology of China, Hefei 230026, China; 4. Institute of Advanced Technology, University of Science and Technology of China, Hefei 230088, China)

Abstract and Keywords

Abstract: Based on the short-stroke method of principal component analysis (PCA) and K-means clustering, this paper constructs the driving conditions of light-duty vehicles in Fuzhou City. Firstly, the original data is preprocessed, the kinematic segments are divided and their characteristic parameters are calculated. Then, PCA is used to reduce the parameter dimension of the divided segments, and K-means cluster analysis is performed to form a typical segment library. According to the minimum parameter deviation method and the maximum correlation coefficient method, the segments are selected to form the driving conditions, and the final driving conditions are output after comparison, and the validity of the final driving conditions is verified. Finally, the Virginia Tech microscopic (VT-Micro) model is used to calculate the single vehicle emission factor combined with the synthetic working conditions, and the motor vehicle emission inventory in Fuzhou City in 2020 is constructed. The study can provide reference for the construction of mobile source emission inventory.

Keywords: driving condition reconstruction; principal component analysis (PCA); K-means clustering; minimum parameter deviation method; maximum correlation coefficient method

基金信息

国家自然科学基金资助项目（61725304；62033012；62103124）；安徽省科技重大专项资助项目（202003a07020009）

合肥工业大学学报(自然科学版)

导航菜单

基于短行程特征聚类的移动源排放清单构建方法