iREbikeLANCE：基于历史使用数据的优化共享单车分配的强化学习代理

IF 2.4 4区计算机科学 Q2 COMPUTER SCIENCE, SOFTWARE ENGINEERING

SoftwareX Pub Date : 2025-10-01 DOI:10.1016/j.softx.2025.102387

Igor Betkier

{"title":"iREbikeLANCE：基于历史使用数据的优化共享单车分配的强化学习代理","authors":"Igor Betkier","doi":"10.1016/j.softx.2025.102387","DOIUrl":null,"url":null,"abstract":"<div><div>This paper introduces iREbikeLANCE, a comprehensive reinforcement learning platform consisting of 2,326 lines of original Python backend code and an interactive web interface. The platform features a custom bike-sharing simulation environment, asynchronous training pipeline, and novel reward engineering interface. iREbikeLANCE integrates real-world BSS data from Warsaw's Veturilo system with a Proximal Policy Optimization (PPO) agent from the Stable Baselines3 library. Its core novelty lies in providing users with an intuitive interface to interactively define and tune multi-component reward function weights for the RL agent. The platform allows users to configure simulation parameters, initiate training, monitor progress via real-time logs and metrics, visualize station states on a map, compare agent performance against baselines, and download resulting models. By abstracting away underlying coding complexity, iREbikeLANCE empowers researchers, students, and practitioners to explore different rebalancing heuristics, understand the impact of reward engineering on agent behavior, and accelerate the development of more efficient and adaptive BSS operations.</div></div>","PeriodicalId":21905,"journal":{"name":"SoftwareX","volume":"32 ","pages":"Article 102387"},"PeriodicalIF":2.4000,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"iREbikeLANCE: Reinforcement learning agent for optimal bike-sharing distribution powered by historical usage data\",\"authors\":\"Igor Betkier\",\"doi\":\"10.1016/j.softx.2025.102387\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>This paper introduces iREbikeLANCE, a comprehensive reinforcement learning platform consisting of 2,326 lines of original Python backend code and an interactive web interface. The platform features a custom bike-sharing simulation environment, asynchronous training pipeline, and novel reward engineering interface. iREbikeLANCE integrates real-world BSS data from Warsaw's Veturilo system with a Proximal Policy Optimization (PPO) agent from the Stable Baselines3 library. Its core novelty lies in providing users with an intuitive interface to interactively define and tune multi-component reward function weights for the RL agent. The platform allows users to configure simulation parameters, initiate training, monitor progress via real-time logs and metrics, visualize station states on a map, compare agent performance against baselines, and download resulting models. By abstracting away underlying coding complexity, iREbikeLANCE empowers researchers, students, and practitioners to explore different rebalancing heuristics, understand the impact of reward engineering on agent behavior, and accelerate the development of more efficient and adaptive BSS operations.</div></div>\",\"PeriodicalId\":21905,\"journal\":{\"name\":\"SoftwareX\",\"volume\":\"32 \",\"pages\":\"Article 102387\"},\"PeriodicalIF\":2.4000,\"publicationDate\":\"2025-10-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"SoftwareX\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S235271102500353X\",\"RegionNum\":4,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"COMPUTER SCIENCE, SOFTWARE ENGINEERING\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"SoftwareX","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S235271102500353X","RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"COMPUTER SCIENCE, SOFTWARE ENGINEERING","Score":null,"Total":0}

引用次数: 0

摘要

iREbikeLANCE是一个综合的强化学习平台，由2326行原始Python后端代码和一个交互式web界面组成。该平台具有定制的共享单车仿真环境、异步训练管道和新颖的奖励工程界面。iREbikeLANCE将来自华沙Veturilo系统的实际BSS数据与来自Stable Baselines3库的近端策略优化（PPO）代理集成在一起。其核心新颖之处在于为用户提供了一个直观的界面来交互式地定义和调整RL agent的多分量奖励函数权重。该平台允许用户配置仿真参数，启动培训，通过实时日志和指标监控进度，在地图上可视化站点状态，将代理性能与基线进行比较，并下载结果模型。通过抽象潜在的编码复杂性，iREbikeLANCE使研究人员、学生和从业者能够探索不同的再平衡启发式方法，了解奖励工程对代理行为的影响，并加速开发更高效和适应性更强的BSS操作。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

iREbikeLANCE: Reinforcement learning agent for optimal bike-sharing distribution powered by historical usage data

This paper introduces iREbikeLANCE, a comprehensive reinforcement learning platform consisting of 2,326 lines of original Python backend code and an interactive web interface. The platform features a custom bike-sharing simulation environment, asynchronous training pipeline, and novel reward engineering interface. iREbikeLANCE integrates real-world BSS data from Warsaw's Veturilo system with a Proximal Policy Optimization (PPO) agent from the Stable Baselines3 library. Its core novelty lies in providing users with an intuitive interface to interactively define and tune multi-component reward function weights for the RL agent. The platform allows users to configure simulation parameters, initiate training, monitor progress via real-time logs and metrics, visualize station states on a map, compare agent performance against baselines, and download resulting models. By abstracting away underlying coding complexity, iREbikeLANCE empowers researchers, students, and practitioners to explore different rebalancing heuristics, understand the impact of reward engineering on agent behavior, and accelerate the development of more efficient and adaptive BSS operations.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

SoftwareX COMPUTER SCIENCE, SOFTWARE ENGINEERING-

CiteScore

5.50

自引率

2.90%

发文量

184

审稿时长

9 weeks

期刊介绍： SoftwareX aims to acknowledge the impact of software on today''s research practice, and on new scientific discoveries in almost all research domains. SoftwareX also aims to stress the importance of the software developers who are, in part, responsible for this impact. To this end, SoftwareX aims to support publication of research software in such a way that: The software is given a stamp of scientific relevance, and provided with a peer-reviewed recognition of scientific impact; The software developers are given the credits they deserve; The software is citable, allowing traditional metrics of scientific excellence to apply; The academic career paths of software developers are supported rather than hindered; The software is publicly available for inspection, validation, and re-use. Above all, SoftwareX aims to inform researchers about software applications, tools and libraries with a (proven) potential to impact the process of scientific discovery in various domains. The journal is multidisciplinary and accepts submissions from within and across subject domains such as those represented within the broad thematic areas below: Mathematical and Physical Sciences; Environmental Sciences; Medical and Biological Sciences; Humanities, Arts and Social Sciences. Originating from these broad thematic areas, the journal also welcomes submissions of software that works in cross cutting thematic areas, such as citizen science, cybersecurity, digital economy, energy, global resource stewardship, health and wellbeing, etcetera. SoftwareX specifically aims to accept submissions representing domain-independent software that may impact more than one research domain.