工業技術研究院 資訊與通訊研究所 許翼麟 張仕穎
前言
在台灣2050淨零轉型策略政策推動之下,預計到2040年電動車銷售市場占比將達到100%。隨著電動車市場的擴大將帶來更高的充電需求,因此政府積極規劃電動車充電站的建置,預計到2025年將建置7,800座公共充電樁,以配合日漸增長的電動車充電需求。為了提升充電站的營運效率,本文將介紹建置前的充電站的選址及案場配置優化、建置後的充電排程最佳化及充電需求管理。此外,介紹由工研院開發的充電站案場營運模擬平台,利用實際數據建模及機器學習技術進行模擬分析,以達到最大化營運商收益和滿足電動車車主的充電需求目的,並提 出優於現行充電排程技術的線性規劃搭配強化學習演算法。希望這些技術能夠協助電動車充電站的營運實現持續穩定獲利,進而吸引更多的民間廠商投入電動車充電站市場,並推動國內電動車產業發展。
精彩內容
1. 電動車充電站營運市場商機 2. 建置前技術:充電站選址及案場配置優化 3. 建置後技術:案場充電排程最佳化及充電需求管理 |
電動車充電站營運市場商機
隨著全球對於環境問題的重視,電動車成為解決汽車行業環保問題的主要選擇之一。國內方面,政府積極推動電動車的發展,並將電動車充電站的建置作為其中一個主要方針。根據政府的台灣2050淨零排放路程徑及策略中的運具電動化策略目標,除2030年市區公車全面電動化外,電動小客車新車年銷售量占所有小客車年銷售量之比例於2030年、2035年與2040年分別達30%、60%與100%,如圖1所示,這項政策也為電動車充電站營運帶來了前所未有的商機。
圖1 FIDO work (資料來源:FIDO Technical Webinar)
現今台灣的電動車市場仍屬於成長期,隨著電動車的普及程度提高,帶動電動車充電需求,電動車充電站的營運商機也與之相對增長。根據工研院產科中心估計,到2025年台灣電動車數量將達到16.4萬輛,而依照歐盟的電動車數量相對於充電樁數量比例為10比1來計算,台灣2025年充電樁需求將為1.64萬樁,電動車充電站營運市場未來可期。
建置前技術:充電站選址及案場配置優化
充電站選址(Siting)
1. 充電需求預測及選址技術
充電站選址首先要做好充電需求的預測,再利用選址最佳化技術從候選位置中找出最佳的布建位置。充電站選址問題如圖2所示,紅色星星代表電動車位置,綠圈黑點代表充電站候選位置,綠色線條為主要道路,透過充電需求預測技術計算出充電需求熱度圖由低至高以藍至紅表示,再透過充電站選址技術決定出最終的充電站選址位置。充電需求預測技術有利用充電站候選位址的周圍興趣點(Point of Interest,POI)及歷史充電站充電資訊並透過深度分群模型(Deep Clustering Model)進行需求預測[2],或結合電網數據及交通網路資料透過圖卷積網路(Graph Convolution Network,GCN)進行需求預測[3],或是根據人口分布、住商利用、家用充電樁數等因素來預測充電需求[4]。選址最佳化技術則包含基因演算法(Genetic Algorithm,GA)[4]和線性規劃(Linear Programming)[5],如考慮多家不同充電級別的電動車充電站營運商,則由賽局理論(Game Theory)模型來找出最佳的充電站布建位置[3][6]。
圖2 充電站需求預測及選址示意圖[6]
2. 充電站選址工具
充電站配置工具比較如表1所列,有美國國家可再生能源實驗室開發的電動車基礎設施投影(Electric Vehicle Infrastructure – Projection,EVI-PRO)工具,和日本電力中央研究所開發的充電站最適配置工具EV-OLYENTOR,並考量了社會經濟統計(如人口密度、人口成長、人民所得等)、車輛規格、充電站密度與距離、交通量/熱區等因素,來做為充電需求預測及充電站布建位置的規劃。
表1 電動車充電站選址工具比較表
|
美國EVI-Pro (2017)
|
日本EV-OLYENTOR
(2012、2014)
|
資料
|
人口資料、氣象資訊、車輛GPS 位置資訊、車輛規格
|
交通部資訊、人口資料、道路車流資訊、車輛GPS位置資訊
|
技術
|
數值模擬
|
數值模擬
|
輸出
|
•充電站需求數目規劃
•充電樁使用率與案場負載變化
|
•公共充電站需求數目規劃
•公共充電站最佳位置分布
|
考量因子
|
社會經濟統計
(人口密度/成長/所得等)
|
V
|
V
|
車輛規格
|
V
|
|
充電站密度與距離
|
V
|
V
|
交通量/熱區
|
V
|
V
|
3. 工研院電動車充電需求預測工具
上述兩項工具都有使用政府收集來的車輛GPS位置資訊作為電動車充電需求的參考因素,然而我國尚未有相關數據資料,因此本文利用交通部高速公路局資料庫中的油車的車次和里程數據轉換為電動車的充電次數和需求電量,如圖3所示。
圖3 工研院電動車充電需求預測工具之運作
案場配置優化(Sizing)
充電站位置選定之後則是要決定案場的供給配置,如充電樁槍數、功率上限、市電契約容量、儲能設備等。除了考量到充電需求(如每個時段的車流、充電需求量等),也要衡量建置成本和服務體驗指標(如電動車車主的等待時間、服務率、滿意度等)以做出最佳的案場配置,如圖4所示。案場配置優化的最佳化技術有考量電動車抵達時間分布、每日車流量、充電時間、充電等待時間、電網上限等因素,利用蒙地卡羅方法(Monde Carlo Method)進行模擬,決定出最好的樁槍數量和儲能裝置大小配置[7]。
圖4 案場配置優化示意
工研院充電站案場配置優化營運模擬平台
工研院目前已開發出一套電動車充電站案場配置優化營運模擬平台,其中分為車流模組、充電站模組。車流模組可設置不同星期和時間段的車流數量、車種、車輛充電模型(Charging Profile)、起始/目標電量狀態、等待時間等參數;充電站模組則可設置樁槍數量、功率上限、儲能設備規格、收費價格、電價方案等參數,電價方案依照台灣電力公司電費表有「非時間電價」、「二段式時間電價」、「三段式時間電價」、「電動車充換電設施電價」等。設定系統參數完後即可進行模擬產生負載情況、服務率、電力成本等資訊,協助充電站營運商能夠找出最佳的案場配置方案。如圖5所示,透過調整樁數、槍數、輸出功率可找出最佳服務率的配置,抑或是調整電價方案來找出最佳的電價方案選擇。
圖5 工研院電動車充電站案場配置優化營運模擬平台
建置後技術:案場充電排程最佳化及充電需求管理
案場充電排程最佳化(Scheduling)
當充電站建置完成後,能夠增加收益的方式之一就是排程調度,將充電需求和充電樁資源進行有效的分配,以最大限度地提高充電站運作效率和車主的滿意度。目前充電排程最佳化問題考量的因素包含不同電力來源(如再生能源、案場儲能、家用儲能)、電價方案(非時間電價、時間電價、動態電價)、最佳化目標(營收、成本、公共利益)等,相關文獻[8-27]整理如圖6所示。充電排程最佳化方法以線性規劃和強化學習演算法(Reinforcement Learning,RL)為主。
圖6 充電站排程最佳化問題文獻整理樹狀圖
工研院案場充電排程最佳化技術
以上文獻之案場規格、電價方案和電動車種都與台灣實際環境有所不同,對演算法的表現有所影響。本文利用上述的充電站營運模擬平台來產生訓練資料,並利用整數線性規劃最佳化(Mixed-Integer Linear Programming,MILP)搭配強化學習方法(Reinforcement Learning,RL) 來決定充電樁及儲能電池的充放電功率。並與兩種傳統演算法做比較,第一種演算法為先到先服務(First Come First serve,FCFS)演算法,即充電站會優先以最大可輸出功率服務先插上充電樁的電動車輛,第二種演算法則是功率共享 (Power Sharing ) 演算法,即充電站會將輸出功率平均分配給充電樁上的電動車輛們,實驗模擬結果相較於兩種演算法都有提升約20%的收益,如圖7所示。
圖7 工研院案場充電排程最佳化技術
充電需求管理(Guiding/Routing)
充電站導流(Guiding)
充電站導流為針對一般電動車主的充電需求提供充電站點推薦,不單只是推薦距離最近的空樁來提升利用率與降低充電焦慮,需考慮交通狀況、路徑能耗、案場即時充電負載等來找出最佳的充電站。充電站導流情境如圖8所示[29],當編號為1、27、33的電動車駕駛(Driver)在分別前往位在節點40、9、9目標充電站的路上抵達交叉路口節點28、16、15時,因考慮即時的道路交通雍塞狀況和位在節點9的充電站使用率過高,將分別目標充電站重新規劃導流到位在節點40、40、9的充電站。充電站導流最佳化技術有結合圖卷積網路(Graph Convolutional Network,GCN)和閘門循環單元(Gated Recurrent Unit,GRU)來預測未來一小段時間內的車流和充電站使用率[28],還有利用分散式共識演算法(distributed consensus algorithm)[29]、結合圖注意力網路(Graph Attention Network,GAT)和深度強化強習(Deep Reinforcement Learning,DRL)方法[30]來模擬充電行為以找出最佳充電站選擇策略。
圖8 充電站導流示意圖[29]
工研院充電站導流技術研發規劃
工研院充電站導流技術研發規劃將與充電站營運商合作,利用目前已開發的充電站模擬平台和案場充電排程最佳化演算法,與開發中的充電需求預測、路徑能耗預測、車輛能耗預測模組結合,協助廠商開發充電站導流演算法並實際使用在用戶手機應用軟體中,提高收益及使用率,同時提供給電動車車主更佳的充電服務體驗,相關研發規劃路線如圖9所示。
圖9 充電站導流示意圖[29]
途程規劃(Routing)
途程規劃為針對電動車物流業者的貨運需求提供完整送取貨點及中繼充電站點的路線規劃,如圖9所示。路徑能耗除了車輛特性外,也會受到上限坡、速度、載重、電池老化、及駕駛行為等因素影響。最佳化方法有線性規劃[31]、基因演算法(Genetic Algorithm,GA)[32][35]、貪婪演算法[33]、蟻群演算法(Ant Colony Optimization,ACO)[34]、多變鄰域搜索算法(Variable Neighborhood Search,VNS)等。
圖10 充電站導流示意圖[29]
工研院途程規劃技術研發規劃
工研院途程規劃技術研發規劃將與物流業者或電巴業者合作,利用目前已開發的充電站模擬平台和案場充電排程最佳化演算法,與開發中的路徑能耗預測、車輛能耗預測模組結合,協助廠商開發途程規劃最佳化技術,降低電力能耗成本,相關研發規劃路線如圖10所示。
結論
除了在充電站選址、案場配置、充電排程及充電需求管理等方面進行最佳化來增進營運收益,充電站營運商還可以利用充電站的充電樁和儲能設備做額外的加值服務,比如結合電動車充電安全檢測技術與充電樁設備,可以為來充電站充電的電動車做電池健康度等安全檢測,做到「充電即安檢」的加值服務。如果充電站有布建儲能設備的話,還能夠在充電需求較低的時段參與台灣電力公司電力交易平台的電力服務市場,達到更有效率的能源利用。
隨著台灣政府積極推動電動車的普及化,充電站營運成為一個重要的市場商機。從充電站選址到案場配置、再到充電排程和充電需求管理,相關最佳化技術的應用有助於提高充電站的運營效率和收益,並滿足車主對於充電的需求,進而促進電動車市場的發展。工研院開發的技術及工具為充電站的運營管理提供了更加精確的解決方案,為未來掌握市場競爭力的重要利器。
參考文獻
[1] (2022) 國家發展委員會-台灣2050淨零排放路徑及策略總說明. Available at: https://www.ndc.gov.tw/Content_List.aspx?n=DEE68AAD8B38BD76
[2] J. Ma, S. Shuai, C. Ji, F. Yin, S. Yuan and T. Yu, "Research on Power Consumption Prediction of Electric Vehicle Charging Station," 2022 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia), Shanghai, China, 2022, pp. 854-858, doi: 10.1109/ICPSAsia55496.2022.9949905.
[3] C. Li, Z. Dong, G. Chen, B. Zhou, J. Zhang and X. Yu, "Data-Driven Planning of Electric Vehicle Charging Infrastructure: A Case Study of Sydney, Australia," in IEEE Transactions on Smart Grid, vol. 12, no. 4, pp. 3289-3304, July 2021, doi: 10.1109/TSG.2021.3054763.
[4] L. Adenaw and M. Lienkamp, "A Model for the Data-based Analysis and Design of Urban Public Charging Infrastructure," 2020 Fifteenth International Conference on Ecological Vehicles and Renewable Energies (EVER), Monte-Carlo, Monaco, 2020, pp. 1-14, doi: 10.1109/EVER48776.2020.9243147.
[5] H. Zhang, S. J. Moura, Z. Hu, W. Qi and Y. Song, "A Second-Order Cone Programming Model for Planning PEV Fast-Charging Stations," in IEEE Transactions on Power Systems, vol. 33, no. 3, pp. 2763-2777, May 2018, doi: 10.1109/TPWRS.2017.2754940.
[6] C. Luo, Y. -F. Huang and V. Gupta, "Placement of EV Charging Stations—Balancing Benefits Among Multiple Entities," in IEEE Transactions on Smart Grid, vol. 8, no. 2, pp. 759-768, March 2017, doi: 10.1109/TSG.2015.2508740.
[7] T. S. Bryden, G. Hilton, B. Dimitrov, C. Ponce de León and A. Cruden, "Rating a Stationary Energy Storage System Within a Fast Electric Vehicle Charging Station Considering User Waiting Times," in IEEE Transactions on Transportation Electrification, vol. 5, no. 4, pp. 879-889, Dec. 2019, doi: 10.1109/TTE.2019.2910401.
[8] W. Tang, S. Bi and Y. J. Zhang, "Online coordinated charging decision algorithm for electric vehicles without future information," in IEEE Transactions on Smart Grid, vol. 5, no. 6, pp. 2810-2824, Nov. 2014, doi: 10.1109/TSG.2014.2346925.
[9] W. Tang and Y. J. Zhang, "A Model Predictive Control Approach for Low-Complexity Electric Vehicle Charging Scheduling: Optimality and Scalability," in IEEE Transactions on Power Systems, vol. 32, no. 2, pp. 1050-1063, March 2017 , doi: 10.1109/TPWRS.2016.2585202.
[10] L. Yao, W. H. Lim and T. S. Tsai, "A Real-Time Charging Scheme for Demand Response in Electric Vehicle Parking Station," in IEEE Transactions on Smart Grid, vol. 8, no. 1, pp. 52-62, Jan. 2017, doi: 10.1109/TSG.2016.2582749.
[11] C. Jin, J. Tang and P. Ghosh, "Optimizing Electric Vehicle Charging: A Customer's Perspective," in IEEE Transactions on Vehicular Technology, vol. 62, no. 7, pp. 2919-2927, Sept. 2013, doi: 10.1109/TVT.2013.2251023.
[12] B. Alinia, M. H. Hajiesmaili, Z. J. Lee, N. Crespi and E. Mallada, "Online EV Scheduling Algorithms for Adaptive Charging Networks with Global Peak Constraints," in IEEE Transactions on Sustainable Computing, vol. 7, no. 3, pp. 537-548, 1 July-Sept. 2022, doi: 10.1109/TSUSC.2020.2979854.
[13] B. Wang and J. Yang, "Optimal electric vehicle charging scheduling with time-varying profits," 2018 52nd Annual Conference on Information Sciences and Systems (CISS), Princeton, NJ, USA, 2018, pp. 1-6, doi: 10.1109/CISS.2018.8362285.
[14] S. Wang, S. Bi and Y. A. Zhang, "Reinforcement Learning for Real-Time Pricing and Scheduling Control in EV Charging Stations," in IEEE Transactions on Industrial Informatics, vol. 17, no. 2, pp. 849-859, Feb. 2021, doi: 10.1109/TII.2019.2950809.
[15] Z. Wei, Y. Li, Y. Zhang and L. Cai, "Intelligent Parking Garage EV Charging Scheduling Considering Battery Charging Characteristic," in IEEE Transactions on Industrial Electronics, vol. 65, no. 3, pp. 2806-2816, March 2018, doi: 10.1109/TIE.2017.2740834.
[16] B. Alinia, M. H. Hajiesmaili and N. Crespi, "Online EV Charging Scheduling With On-Arrival Commitment," in IEEE Transactions on Intelligent Transportation Systems, vol. 20, no. 12, pp. 4524-4537, Dec. 2019, doi: 10.1109/TITS.2018.2887194.
[17] Q. Chen et al., "Dynamic Price Vector Formation Model-Based Automatic Demand Response Strategy for PV-Assisted EV Charging Stations," in IEEE Transactions on Smart Grid, vol. 8, no. 6, pp. 2903-2915, Nov. 2017, doi: 10.1109/TSG.2017.2693121.
[18] Q. Yan, B. Zhang and M. Kezunovic, "Optimized Operational Cost Reduction for an EV Charging Station Integrated With Battery Energy Storage and PV Generation," in IEEE Transactions on Smart Grid, vol. 10, no. 2, pp. 2096-2106, March 2019, doi: 10.1109/TSG.2017.2788440.
[19] Y. Zhou, D. K. Y. Yau, P. You and P. Cheng, "Optimal-Cost Scheduling of Electrical Vehicle Charging Under Uncertainty," in IEEE Transactions on Smart Grid, vol. 9, no. 5, pp. 4547-4554, Sept. 2018, doi: 10.1109/TSG.2017.2662801.
[20] Y. -T. Liao and C. -N. Lu, "Dispatch of EV Charging Station Energy Resources for Sustainable Mobility," in IEEE Transactions on Transportation Electrification, vol. 1, no. 1, pp. 86-93, June 2015, doi: 10.1109/TTE.2015.2430287.
[21] Y. Yuan, L. Jiao, K. Zhu and L. Zhang, "Scheduling Online EV Charging Demand Response via V2V Auctions and Local Generation," in IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 8, pp. 11436-11452, Aug. 2022, doi: 10.1109/TITS.2021.3103970.
[22] J. Jin and Y. Xu, "Optimal Policy Characterization Enhanced Actor-Critic Approach for Electric Vehicle Charging Scheduling in a Power Distribution Network," in IEEE Transactions on Smart Grid, vol. 12, no. 2, pp. 1416-1428, March 2021, doi: 10.1109/TSG.2020.3028470.
[23] Z. Wan, H. Li, H. He and D. Prokhorov, "Model-Free Real-Time EV Charging Scheduling Based on Deep Reinforcement Learning," in IEEE Transactions on Smart Grid, vol. 10, no. 5, pp. 5246-5257, Sept. 2019, doi: 10.1109/TSG.2018.2879572.
[24] B. Upadhaya, D. Feng, Y. Zhou, Q. Gui, X. Zhao and D. Wu, "Optimal decision making model of battery energy storage-assisted electric vehicle charging station considering incentive demand response," 8th Renewable Power Generation Conference (RPG 2019), Shanghai, China, 2019, pp. 1-8, doi: 10.1049/cp.2019.0499.
[25] T. Long, Q. -S. Jia, G. Wang and Y. Yang, "Efficient Real-Time EV Charging Scheduling via Ordinal Optimization," in IEEE Transactions on Smart Grid, vol. 12, no. 5, pp. 4029-4038, Sept. 2021, doi: 10.1109/TSG.2021.3078445.
[26] T. Zeng, S. Bae, B. Travacca and S. Moura, "Inducing Human Behavior to Maximize Operation Performance at PEV Charging Station," in IEEE Transactions on Smart Grid, vol. 12, no. 4, pp. 3353-3363, July 2021, doi: 10.1109/TSG.2021.3066998.
[27] L. Yan, X. Chen, J. Zhou, Y. Chen and J. Wen, "Deep Reinforcement Learning for Continuous Electric Vehicles Charging Control With Dynamic User Behaviors," in IEEE Transactions on Smart Grid, vol. 12, no. 6, pp. 5124-5134, Nov. 2021, doi: 10.1109/TSG.2021.3098298.
[28] S. Su, Y. Li, Q. Chen, M. Xia, K. Yamashita and J. Jurasz, "Operating Status Prediction Model at EV Charging Stations With Fusing Spatiotemporal Graph Convolutional Network," in IEEE Transactions on Transportation Electrification, vol. 9, no. 1, pp. 114-129, March 2023, doi: 10.1109/TTE.2022.3192285.
[29] X. Shi, Y. Xu, Q. Guo, H. Sun and W. Gu, "A Distributed EV Navigation Strategy Considering the Interaction Between Power System and Traffic Network," in IEEE Transactions on Smart Grid, vol. 11, no. 4, pp. 3545-3557, July 2020, doi: 10.1109/TSG.2020.2965568.
[30] Q. Xing, Y. Xu and Z. Chen, "A Bilevel Graph Reinforcement Learning Method for Electric Vehicle Fleet Charging Guidance," in IEEE Transactions on Smart Grid, doi: 10.1109/TSG.2023.3240580.
[31] J. Lin, W. Zhou and O. Wolfson, "Electric vehicle routing problem," in Transportation Research Procedia, vol. 12, pp. 508-521, 2016, doi: 10.1016/j.trpro.2016.02.007.
[32] H. Yang, S. Yang, Y. Xu, E. Cao, M. Lai and Z. Dong, "Electric Vehicle Route Optimization Considering Time-of-Use Electricity Price by Learnable Partheno-Genetic Algorithm," in IEEE Transactions on Smart Grid, vol. 6, no. 2, pp. 657-666, March 2015, doi: 10.1109/TSG.2014.2382684.
[33] M. Granada-Echeverri, L. C. Cubides and J. O. Bustamante, "The Electric Vehicle Routing Problem with Backhauls," in International Journal of Industrial Engineering Computations, vol. 11, no. 1, pp. 131-152, 2020, doi: 10.5267/j.ijiec.2019.6.001.
[34] Y. -H. Jia, Y. Mei and M. Zhang, "A Bilevel Ant Colony Optimization Algorithm for Capacitated Electric Vehicle Routing Problem," in IEEE Transactions on Cybernetics, vol. 52, no. 10, pp. 10855-10868, Oct. 2022, doi: 10.1109/TCYB.2021.3069942.
[35] J. Xiao, et al. "A diversity-enhanced memetic algorithm for solving electric vehicle routing problems with time windows and mixed backhauls," in Applied Soft Computing, vol. 134, Feb. 2023, doi: 10.1016/j.asoc.2023.110025.