{"title":"用于 xEV 充电推荐系统的物理信息冷启动能力","authors":"Raik Orbay;Aditya Pratap Singh;Johannes Emilsson;Michele Becciani;Evelina Wikner;Victor Gustafson;Torbjörn Thiringer","doi":"10.1109/OJVT.2024.3469577","DOIUrl":null,"url":null,"abstract":"An effortless charging experience will boost electric vehicle (xEV) adoption and assure driver satisfaction. Tailoring the charging experience incorporating smart algorithms introduces an exciting set of development opportunities. The goal of a smart charging algorithm is to lay down an accurate estimation of charging power needs for each user. As recommender systems (RS) are frequently used for tailored services and products, a novel RS based approach is developed in this study. Based on a collaborative-filtering principle, an RS agent will customize charging power transient prioritizing the physical principles governing the battery system, correlated to customer preferences. However, parallel to other RS applications, a collaborative-filtering for charging power transient design may suffer from the cold-start problem. This paper thus aims to prescribe a remedy for the cold-start problem encountered in RS specifically for charging power transient design. The RS is cold-started based on multiphysical modelling, combined with customer driving styles. It is shown that using 7 fundamental charging power transients would capture about 70% of a set of representative charging power transient population. Matching a unsupervised learning based clustering pipeline for 7 possible customer driving styles, an RS agent can prescribe 7 charging power transients automatically and cold-start the RS until more data is available.","PeriodicalId":34270,"journal":{"name":"IEEE Open Journal of Vehicular Technology","volume":"5 ","pages":"1457-1469"},"PeriodicalIF":5.3000,"publicationDate":"2024-09-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10697286","citationCount":"0","resultStr":"{\"title\":\"A Physics-Informed Cold-Start Capability for xEV Charging Recommender System\",\"authors\":\"Raik Orbay;Aditya Pratap Singh;Johannes Emilsson;Michele Becciani;Evelina Wikner;Victor Gustafson;Torbjörn Thiringer\",\"doi\":\"10.1109/OJVT.2024.3469577\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"An effortless charging experience will boost electric vehicle (xEV) adoption and assure driver satisfaction. Tailoring the charging experience incorporating smart algorithms introduces an exciting set of development opportunities. The goal of a smart charging algorithm is to lay down an accurate estimation of charging power needs for each user. As recommender systems (RS) are frequently used for tailored services and products, a novel RS based approach is developed in this study. Based on a collaborative-filtering principle, an RS agent will customize charging power transient prioritizing the physical principles governing the battery system, correlated to customer preferences. However, parallel to other RS applications, a collaborative-filtering for charging power transient design may suffer from the cold-start problem. This paper thus aims to prescribe a remedy for the cold-start problem encountered in RS specifically for charging power transient design. The RS is cold-started based on multiphysical modelling, combined with customer driving styles. It is shown that using 7 fundamental charging power transients would capture about 70% of a set of representative charging power transient population. Matching a unsupervised learning based clustering pipeline for 7 possible customer driving styles, an RS agent can prescribe 7 charging power transients automatically and cold-start the RS until more data is available.\",\"PeriodicalId\":34270,\"journal\":{\"name\":\"IEEE Open Journal of Vehicular Technology\",\"volume\":\"5 \",\"pages\":\"1457-1469\"},\"PeriodicalIF\":5.3000,\"publicationDate\":\"2024-09-27\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10697286\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"IEEE Open Journal of Vehicular Technology\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://ieeexplore.ieee.org/document/10697286/\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Open Journal of Vehicular Technology","FirstCategoryId":"1085","ListUrlMain":"https://ieeexplore.ieee.org/document/10697286/","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
A Physics-Informed Cold-Start Capability for xEV Charging Recommender System
An effortless charging experience will boost electric vehicle (xEV) adoption and assure driver satisfaction. Tailoring the charging experience incorporating smart algorithms introduces an exciting set of development opportunities. The goal of a smart charging algorithm is to lay down an accurate estimation of charging power needs for each user. As recommender systems (RS) are frequently used for tailored services and products, a novel RS based approach is developed in this study. Based on a collaborative-filtering principle, an RS agent will customize charging power transient prioritizing the physical principles governing the battery system, correlated to customer preferences. However, parallel to other RS applications, a collaborative-filtering for charging power transient design may suffer from the cold-start problem. This paper thus aims to prescribe a remedy for the cold-start problem encountered in RS specifically for charging power transient design. The RS is cold-started based on multiphysical modelling, combined with customer driving styles. It is shown that using 7 fundamental charging power transients would capture about 70% of a set of representative charging power transient population. Matching a unsupervised learning based clustering pipeline for 7 possible customer driving styles, an RS agent can prescribe 7 charging power transients automatically and cold-start the RS until more data is available.