設計用於可靠電動車充電通信系統的速率自適應方法

Abstract

隨著電動汽車在全球普及，充電的問題也越來越受到重視。因此，衍生出了多種充電通訊標準，例如：CHAdeMO，ISO 15118 等其他充電通訊標準。其中，最新推出的 ISO 15118-20 標準相較於其他充電通訊標準新增了包括無缐充電(WPT)，雙向充電 (BPT) 等新功能，提供給電動車用戶更豐富的充電場景。本研究中，實作了 ISO 15118-8 無線充電底層標準和部分 ISO 15118-20 充電上層標準。其中，ISO 15118-20標準中規定了每個通訊階段的超時時間。頻繁的超時行為會嚴重影響整個充電系統的效能。因此，爲了盡量減少充電系統在真實環境下發生超時導致通訊中斷，我們在實作的系統上對每一個通訊階段進行延遲量測。爲此，我們還設計了兩種現實中可能出現的情境：1) 多個用戶共享一個充電 AP ; 2) 用於充電的 AP 周圍有相同通道的干擾 AP。基於上述兩種情境，我們的實驗結果表明：在第一種情境下，隨著共存用戶數量的增加，通訊延遲也相應增加，如果共存用戶超過一定數量，則會導致超時頻繁發生。此外，我們還發現在共存用戶數相同時，5GHz 完成充電過程的總延遲顯著低於 2.4GHz 下的延遲。在第二種情境下，我們觀察到由於相同通道干擾 AP 的加入，導致通訊延遲增加和超時發生。爲了避免由於干擾 AP 導致的超時，我們設計了一種 802.11 速率調配演算法，并將該算法實作到 AP 中。經實驗證明，我們設計的演算法可以有效降低由於干擾AP導致的超時發生。

With the increasing popularity of electric vehicles (EVs) worldwide, EV charging has become a matter of increasing concern. This has led to the emergence of various charging communication standards such as CHAdeMO, ISO 15118, among others. The latest ISO 15118-20 standard has introduced new features like wireless power transfer (WPT) and bidirectional power transfer (BPT), offering EV users a wider range of charging scenar ios compared to other standards. In this study, we implemented the lower layer standard of ISO 15118-8 wireless charging and part of the upper layer standard of ISO 15118-20 charging. The ISO 15118-20 standard outlines the timeout period for each communication phase. Frequent timeouts can severely impact the performance of the entire charging sys tem. Hence, to minimize communication interruptions caused by timeouts in real-world charging systems, we measured the latency for each communication stage in our implemented system. We designed two realistic scenarios: 1) multiple users sharing a charging access point (AP); and 2) interference from APs using the same channel around the AP designated for charging. Based on these scenarios, our experimental results demonstrate that in the first scenario, as the number of concurrent users increases, communication la tency correspondingly rises, and exceeding a certain threshold of users can lead to frequent timeouts. Moreover, we found that the total latency for completing the charging process is considerably lower in the 5GHz communication scenario than in the 2.4GHz scenario. In the second scenario, we observed that the addition of interfering APs on the same channel resulted in increased communication latency and timeouts. To mitigate timeouts caused by interfering APs, we developed an 802.11 rate adaptation algorithm and implemented it in the AP. The experimental results revealed that our designed algorithm can effectively reduce timeouts caused by interfering APs.