隨著地球能源的減少,油耗的課題是日益重要。本研究與客運公司合作,以數位行車記錄器收集行車時每秒的速度、加速度、經緯度等其他相關的數位資訊,探討司機的駕駛行為。另一方面,以人格特質問卷方式來收集資料,使用因素分析與集群分析對司機駕駛分群。最後結合數位資訊的駕駛行為以及問卷資料的駕駛分群作為預測因子運用迴歸分析建立影響油耗的因子及模型。 本研究主要針對三個問題來探討,其一為尋找重要的油耗影響因子,其二是上述的重要因子其影響油耗的數據表現為何, 最後是探討除了駕駛行為外,不同類型的司機是否也會影響油耗的表現。 結果顯示,在不同時速下的加速與煞車表現,其影響油耗程度不一,平穩因子則具有省油表現,可值得繼續後續研究探討。而不同類型的駕駛在油耗上的表現,依現有油價換算成金額,一天便可有200元的差異,長期下來差距相當可觀,可提供給業者在徵才上需多加考量與注意的地方。
As the Earth's energy reduction, fuel consumption issue becomes more important day by day. In this study, we cooperated with passenger transport company to collect digital driving information such as speed per hour, acceleration, latitude and longitude, etc. every second by tachograph equipment to analysis driver's driving patterns. On the other hand, we applied factor analysis and cluster analysis to categorize drivers into three different types by their personality through driving habits questionnaire. Finally, digital information and the personality of the questionnaire data for driving grouping were integrated to establishing a fuel consumption model with regression analysis. The purpose of this study is the following three issues: finding important factors affecting oil consumption, estimating the impact of these important factors, and distinguishing the effects for different types of drivers. The results showed that the performance of acceleration and brake has different impact on fuel consumption under different speed. The stable factor has a fuel efficient impact, it's a good subject to continue a follow-up study. However, we can convert the different oil consumption of three driving grouping to money value. The difference can be up to 200 NT$ per day. For passenger transport company, that will be a good direction to attention different driving type when interview a new employee.