隨著全球暖化與油價的持續上揚，節能駕駛的議題逐漸受重視。現有節能輔助裝置僅提供即 時行車油耗資訊給駕駛參考，使駕駛修正自身的駕駛方式。但對於經驗不足或習慣不好的駕駛人 而言，並不易立即透過油耗的變化趨勢，來對應與修正自身的駕駛行為。有鑑於此，本論文提出 一基於車輛動態與行車負載估測之智慧型節能駕駛輔助系統，該系統即時擷取車內診斷系統資 訊，再利用函數近似方式建立瞬間油耗模型，接著再以模糊推論估測駕駛節能等級。而為更有效 協助駕駛達到節省油耗之目的，本系統使用模糊規則建立車輛負載模型。最後再以內建人工智慧 即時分析出適當地駕駛節能建議，來提供即時的駕駛操作參考。本系統包含一智慧車輛資料嵬集 器與一搭載Android節能應用程式之智慧型裝置。透過實車安裝與測試，本系統可提升至少7%以 上的行駛油耗。

Eco-driving skill has been getting more and more attentions because of global warming and increasing oil price. So far, existing eco-driving assistance systems mainly offered raw instantaneous fuel economy to driver. However, inexperience driver still had the difficulty to turn raw fuel economy information into proper eco-driving behavior. For this situation, an intelligent eco-driving suggestion system based on vehicle loading model was developed. The instantaneous fuel economy was computed according to the information from vehicle on board diagnostic system. In addition, fuzzy inference system was applied to estimate eco-level and fuzzy rules were utilized to establish a vehicle loading model. The appropriate eco-driving suggestion was analyzed by built-in artificial intelligence and can be displayed on any Android portable device. Finally, the developed eco-driving suggestion system was ported on Smart Vehicle Information Gateway, installed on real vehicle and tested on real track. The experimental results proved that 7% fuel economy can be improved.