ELECTRIC MOTORCYCLE EMISSION ANALYSIS THROUGH COMPUTER VISION
Other Title
利用電腦視覺分析電動機車所造成之空氣汙染
Journal
中國土木水利工程學刊
Journal Volume
34
Journal Issue
7
Start Page
631
End Page
637
ISSN
1015-5856
Date Issued
2022-11
Author(s)
Abstract
The motorcycle is the transport with the highest ownership ratio in Taiwan in advantage of its low cost, high mobility, and parking convenience. The ownership ratio of the motorcycle is about twice that of the passenger vehicle. Pollution emission from motorcycles is an issue of importance in traffic-related environmental problems. In spite of the related literature on pollution emission from internal combustion engine motorcycles, there is little literature on emission from electric motorcycles. As the percentage of registered electric motorcycles in registered motorcycles rises, it is necessary to study the pollutants related to electric motorcycles. This study aims at analyzing non-exhaust emissions from electric motorcycles. A computer vision-based motorcycle classification method and roadside pollutant concentration measurement are utilized to study the pollutants that electric motorcycles could produce in driving process. In this study, we identified coarse particle and PM1.0 as the pollutants related to the emissions from electric motorcycles. This research could be applied for the benefit evaluation of vehicle fleet electrification.
由於機車的低開銷、高機動性及停泊便利性，其為我國持有率最高之運具。機車污染排放是重要的交通環境議題，儘管前人已對內燃機機車排放進行研究，鮮少有針對電動機車排放的研究。車輛排放包含尾氣排放和非尾氣排放，電動機車雖然不會產生尾氣排放，但仍然會產生非尾氣排放。本研究旨在分析電動機車的非尾氣排放，透過以電腦視覺為基礎的機車分類方法，搭配污染物環境濃度監測，定性分析電動機車行駛過程中可能製造的污染物。於此研究中，我們發現粗顆粒物和PM1.0為電動機車可能排放的污染物。這項研究可為機車電氣化的效益評估提供參考。
Subjects
electric two-wheeler
non-exhaust emission
machine learning
deep learning
電動機車
非尾氣排放
機器學習
深度學習
Type
journal article