[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-86212-en":3,"doc-seo-86212-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},86212,1374391974564,"Clementine","https://ap-avatar.wpscdn.com/avatar/14000253aa45c000a9e?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779874745381141002",8,"Research & Report","Inter Stop Energy Prediction and Causal Driver Quantification for Dual Source Trolleybuses via a Time Aware Tabular Deep Learning Architecture","Dual-source trolleybuses switch between overhead catenary power and on-board battery operation, producing inter-stop energy patterns shaped by route attributes, high-frequency vehicle trajectories, and hourly meteorological conditions. Existing methods encode heterogeneous inputs inadequately and rarely provide causal explanations. The proposed time-aware tabular deep learning architecture jointly represents static and sequential features, while a Bayesian optimization and three-layer causal pipeline move analysis from correlation to causal mechanisms. Experiments on the Zurich dataset report 6.52% MAPE and R²=0.982.","Inter-Stop Energy Prediction and Causal Driver Quantification for DualSource Trolleybuses via a Time-Aware Tabular Deep Learning Architecture  \nWentao Zeng a,b , Zijian Huang c, Yiming Bie d, Jiabin Wu a,*, Jun Gong ea School of Management, Foshan University, Foshan 528000, China  \nb School of Mechanical and Electrical Engineering and Automation, Foshan University, Foshan 528225, China  \nc School of Artificial Intelligence, South China Normal University, Guangzhou 510631, China d School of Transportation, Jilin University, Changchun 130022, China  \ne Department of Civil Engineering, The University of HongKong, HongKong 999077, China  \n* Corresponding author: Jiabin Wu (E-mail [address:](address: jiabinwu@fosu.edu.cn)[ j](address: jiabinwu@fosu.edu.cn)[iabinwu@fosu.edu.cn](address: jiabinwu@fosu.edu.cn))  \nAbstract：Dual-source trolleybuses switch between overhead catenary power and on-board battery operation, producing energy patterns shaped by static route attributes, high-frequency vehicle trajectories, and hourly meteorological conditions. Existing models encode such heterogeneous inputs inadequately, and none provides causal explanation of the factors governing energy use. This paper addresses both limitations. On the artificial intelligence side, a time-aware tabular deep learning architecture is proposed. Periodic time encoding is embedded into a parameter-efficient batch-ensemble backbone, enabling static and sequential features to be represented jointly in a single network. Bayesian optimization guided by tree-structured density estimation drives the hyperparameter search. A three-layer causal explanation pipeline then extends the predictive model: featurelevel attribution quantifies marginal contributions, a linear non-Gaussian acyclic model recovers causal directions from observational data, and a meta-learner estimates net average treatment effects—moving the analysis from correlation to causal mechanism. On the engineering side, the framework targets inter-stop energy management for dual-source trolleybuses in urban transit networks. Experiments on the Zurich trolleybus dataset, enriched with meteorological records, yield a mean absolute percentage error of 6.52% and a coefficient of determination of 0.982, surpassing all ten baseline models across statistical, tree-ensemble, and deep learning categories. Ablation experiments confirm periodic time encoding as the principal source of accuracy gain. Causal analysis identifies the regenerative braking ratio and average speed as the dominant energy-saving levers. Coasting distance, by contrast, emerges as the primary contributor to excess consumption. These results provide concrete intervention thresholds for vehicle technology, driving behavior, capacity allocation, and catenary network planning.  \nKeywords: dual-source trolleybus; energy consumption prediction; deep learning; bayesian optimization; interpretability analysis; causal inference  \n1. Introduction  \nTransport has become the largest urban source of air pollution in most cities worldwide, and pressure for energy saving and emission reduction continues to mount [1-3] . The IEA Global EV Outlook 2025 [4] reports that in 2024 the global electric fleet displaced more than 1.3 million barrels of oil per day, and electric bus sales grew by 30% year-on-year to surpass 70,000 units, underscoring the central role of transit electrification in the energy transition. Pure battery-electric technology, however, still faces a structural payload-range-cost dilemma in heavy-duty commercial applications: highly redundant battery packs account for close to 50% of the total purchase cost, which severely constrains operational economics. Although trams, trolleybuses, and batteryelectric buses have all been deployed at scale, reconciling operational efficiency and range flexibility under limited infrastructure investment remains a critical bottleneck for the sector.  \nThe dual-source trolleybus (DTB), with its hybrid supply arch","cbCaieQs5aXBkyrk","https://ap.wps.com/l/cbCaieQs5aXBkyrk","pdf",3435077,2,1,42,"English","en",105,"# Introduction\n## Dual-source trolleybus energy challenges\n# Method\n## Time-aware tabular deep learning architecture\n## Bayesian optimization for hyperparameter search\n## Causal explanation pipeline","[{\"question\":\"What problem does the paper address for dual-source trolleybuses?\",\"answer\":\"It tackles inaccurate energy prediction and the lack of causal explanations for what drives energy use between stops under overhead-catenary and battery switching conditions.\"},{\"question\":\"How does the proposed model represent heterogeneous inputs?\",\"answer\":\"It embeds periodic time encoding into a parameter-efficient batch-ensemble backbone so static route attributes and high-frequency sequential factors are learned together in one network.\"},{\"question\":\"Which factors are found to dominate energy savings and excess consumption?\",\"answer\":\"Causal analysis identifies the regenerative braking ratio and average speed as dominant energy-saving levers, while coasting distance is the primary contributor to excess consumption.\"}]",1784209495,106,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"inter-stop-energy-prediction-and-causal-driver-quantification-for-dual-source-trolleybuses-via-a-time-aware-tabular-deep-learning-architecture","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,47,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":20},"https://docshare.wps.com/document/","Document",{"item":48,"name":12,"@type":43,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/inter-stop-energy-prediction-and-causal-driver-quantification-for-dual-source-trolleybuses-via-a-time-aware-tabular-deep-learning-architecture/86212/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-26","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the paper address for dual-source trolleybuses?","Question",{"text":75,"@type":76},"It tackles inaccurate energy prediction and the lack of causal explanations for what drives energy use between stops under overhead-catenary and battery switching conditions.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the proposed model represent heterogeneous inputs?",{"text":80,"@type":76},"It embeds periodic time encoding into a parameter-efficient batch-ensemble backbone so static route attributes and high-frequency sequential factors are learned together in one network.",{"name":82,"@type":73,"acceptedAnswer":83},"Which factors are found to dominate energy savings and excess consumption?",{"text":84,"@type":76},"Causal analysis identifies the regenerative braking ratio and average speed as dominant energy-saving levers, while coasting distance is the primary contributor to excess consumption.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]