近年來隨著全球暖化的加劇，環保及永續發展的議題持續受到各國政府與企業的重視。儘管城市物流是支持實體經濟以及虛擬經濟的後端基礎，但其中的貨物運輸對環境產生的負面影響甚鉅，如運輸過程中燃料的消耗及溫室氣體的排放等，且貨物運輸造成的交通壅塞及空氣汙染亦會使城市居住品質下降。此外，隨著及時制度（Just-in-time）及電子商務的重要性日益增加，顧客對於貨物送達時間的要求也越趨嚴格。因此，一個能最小化環境影響並滿足顧客時窗（Time windows）要求的綠色物流（Green logistics）規劃模型是必須的。 本研究提出了一個基於車輛途程問題（Vehicle routing problem）的綠色城市物流模型，規劃目標為在滿足顧客的貨物需求及時窗需求的情況之下，最小化所有送貨車輛的碳排放量。模型中除了考量車輛行駛時所產生的碳排放，更考量了車輛在顧客位置等待時所產生的閒置碳排放，使本研究的規劃模型能適用於運送時需進行溫度控管的貨物上。本研究亦針對此模型提出一個啟發式演算法，並帶入包含100個節點的模擬案例中，與傳統僅以行駛距離、行駛時間作為規劃目標的演算法比較，最後得到本研究的演算法在綠色目標、經濟目標上皆優於傳統演算法的結果。;Following the concern of global warming in recent years, the issues of environmental protection and sustainability have been receiving increasing attention from both governments and businesses. City logistics is the foundation of economy; however, its freight transportation has a considerable negative impact on the environment due to fuel consumptions and greenhouse gas emissions. Freight transportation also causes traffic congestion and air pollution, which may lead to a decrease in the quality of life of urban residents. Furthermore, the growing significance of just-in-time production and e-commerce contributes to the need of satisfying the customers’ increasing demand for shorter delivery time windows. Therefore, the necessity of a green logistics planning model that enables the minimization of the environmental impact and facilitates the adherence to these time windows is evident. The present research proposes a green city logistics planning model based on the vehicle routing problem (VRP). The goal is to minimize the carbon emissions of all delivery vehicles, on the condition of fulfilling the customers’ demands for goods and adhering to time windows. The model not only considers the emissions of vehicles in motion (hot emissions) but also the emissions of vehicles under idling (idle emissions). Therefore, this model can be applied to cases where a control of the temperature of goods is required. In addition, this research proposes a heuristic algorithm for the model and uses it to solve the benchmark problems with 100 nodes. Lastly, the solution of the created algorithm is compared with the one of a traditional VRP algorithm. The results of this comparison show that the algorithm developed during the present research performs better in both environmental and economic terms.
