基於時變參數的電動車充電系統最佳途程規劃研究

Abstract

本研究提出了一個以資料驅動的途程策略方法，結合統計管制圖的概念，建構出一個最佳途程策略，旨在最小化從發出充電需求到完成充電所需的整體時間。我們採用了最鄰近K充電站的途程策略（Nearest-K RoutingPolicy），並利用EWMA管制圖來監控車流量的顯著變動，使得在節省計算成本的同時，能靈活因應車流量的變化。研究結果顯示，當輸入流量產生重大變化時，使用最鄰近K充電站的途程策略來調整最佳途程策略，能有效降低系統的平均交通時間和等待時間，提升充電系統的整體效率。整體而言，本研究提出的結合最鄰近K充電站途程策略與EWMA管制圖的方法，為應對隨時間變動的輸入參數，提供了一種高效的電動車充電系統途程策略。

This study proposes a data-driven routing strategy combined with the concept of statistical control charts to construct an optimal routing strategy, aiming to minimize the overall time from initiating a charging request to completing the charge. We adopted the Nearest-K Routing Policy and used the EWMA control chart to monitor significant changes in traffic flow, allowing for flexible adaptation to traffic changes while saving computational costs. The research results show that when there are significant changes in the input flow, adjusting the optimal routing strategy using the Nearest-K Routing Policy can effectively reduce the system’s average travel time and waiting time, enhancing the overall efficiency of the charging system. Overall, the proposed method,which combines the Nearest-K Routing Policy with the EWMA control chart, provides an efficient routing strategy for electric vehicle charging systems with time-varying input parameters.