電動卡車路線與移動式充電服務之雙層最佳化問題

Abstract

隨著全球暖化問題日益嚴重，各國政府紛紛制定嚴格的減碳目標和環境法規，以減少溫室氣體排放。燃油車作為主要排放源之一，正逐漸被電動車取代。近年來，電動車的市場占有率確實逐年增長，成為汽車市場的重要組成部分。而因其具備的長期經濟效益和環境影響減少的優勢，電動車在物流配送領域的應用也日益增加。 然而電動車受到行駛里程的限制，再加上現有的充電基礎設施的不足，物流公司使用電動車進行配送時可能會為了充電而額外增加旅行成本及時間。而新的技術移動式充電車能隨時隨地的移動來為電動車充電，提供更靈活的充電選擇，因此本研究將引入移動式充電車技術於電動卡車物流配送問題中。 根據移動式充電服務的提供者，本研究探討了兩種情境：一為物流公司擁有並自行調度移動式充電車隊；二為物流公司委託專業移動式充電服務公司提供充電服務。在第一種情境下，建立了利用移動式充電之電動卡車配送問題，以最小化營運成本為目標，探討移動式充電車在物流配送中的效益。在第二種情境下，通過建立雙層最佳化模型，深入探討物流公司與移動式充電服務公司之間的合作機制及其成本效益。為了有效解決雙層混合整數規劃問題，本研究使用CCG演算法求解，並通過案例測試以及與傳統迭代方法比較，以驗證所提模型的合理性與實用性。 研究結果顯示引入移動式充電車，能有效降低物流公司營運成本並服務更多顧客，並且讓電動卡車的路徑選擇上更有彈性。而CCG演算法也可以在一定時間內求得較迭代方法更好的解，並且是雙層最佳化的精確解。

With the growing concerns in modern societies regarding greenhouse gas emissions and climate change, many countries have made commitments to reduce emissions. As an alternative to internal combustion vehicles (ICVs), the adoption rate of electric vehicles (EVs) has been increasing annually. Consequently, the use of EVs in the logistics and distribution sector is also on the rise. However, range anxiety, long recharge durations, and insufficient recharging infrastructure still restrain the wider adoption of EVs in the sector. To address these issues, the new technology of mobile charging stations (MCS) can provide flexible charging options by moving to charge EVs anytime and anywhere. Therefore, this study introduces MCS technology into the Electric Vehicle Routing Problem with Time Windows (EVRPTW). Based on the provision of mobile charging services, this study explores two scenarios: one where logistics companies own and operate their own fleet of mobile charging stations, and another where they outsource charging services to professional MCS companies. In the first scenario, we develop a model for EVRPTW utilizing mobile charging stations with the objective of minimizing operational costs, thereby examining the benefits of MCS in logistics distribution. In the second scenario, we establish a bilevel optimization model to deeply investigate the cooperation mechanisms and cost efficiencies between logistics companies and MCS companies. To effectively solve the bilevel mixed-integer programming problem, this study employs iterative methods and develops a Column-and-Constraint Generation Algorithm. Case studies are conducted to validate the proposed model's reasonableness and practicality. The results show that the introduction of mobile charging stations can significantly reduce logistics companies' operational costs and serve more customers while providing greater flexibility in electric vehicle routing. Additionally, the Column-and-Constraint Generation Algorithm can achieve better solutions within a reasonable time compared to iterative methods, providing exact solutions for bilevel optimization.