NextGen Data-Driven Timetable Performance Optimisation Tool

下一代数据驱动的时间表性能优化工具

• 批准号: 10037862
• 金额: $ 20.11万
• 依托单位: JNCTION LIMITED
• 依托单位国家: 英国
• 项目类别: Collaborative R&D
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=10037862
• 关键词: NextGen; Data; Driven; Timetable; Performance

During the pandemic, the on-time reliability of services significantly increased due to the reduction in the number of services and passengers.However, as passengers have returned to the railway performance has once again deteriorated. This has an even greater impact on the industry post-pandemic as passengers' expectations for services that are reliable and run on-time is even higher. Increased delays and passenger dissatisfaction therefore leads to an even greater decreased revenue from ticket sales.Poor performance is in large part due to a poorly planned timetable that is often operationally unachievable or cannot handle minor perturbations. This is due to the timetable usually being planned with simulations and the method does not in how trains performing in reality at junction or stations.Through years of working closely with performance, planning and operational teams, we've identified that by using granular train movement data and machine learning techniques, the actual performance of the existing timetable could be accurately calculated. This would enable planners with accurate information to make faster and better planning decisions that are based on real-world evidence.Our Timetable Analysis tool will deliver automatically updated insights and recommendations to planners that is highly aligned to the planning process. Utilising both on-track (track circuit) and on-fleet (GPS and OTMR) data, the tool will provide an integrated view to both Network Rail and TOC teams.Fundamentally this tool will result in a step change in the speed and quality of timetable planning, moving away from the use of limited simulations and anecdotal experience to a fully evidenced-based approach.

疫情期间，由于班次和乘客数量的减少，服务的准点率大幅提高。然而，随着乘客返回铁路，服务表现再次恶化。这对疫情后的行业产生了更大的影响，因为乘客对可靠和准时运行的服务的期望更高。因此，更多的延误和乘客的不满导致售票收入更大的下降。糟糕的表现在很大程度上是由于计划不周的时间表，在操作上往往无法实现或无法处理轻微的干扰。这是因为列车时刻表通常是通过模拟来规划的，而该方法并不能反映列车在路口或车站的实际表现。通过与性能、规划和运营团队多年的密切合作，我们已经发现，通过使用细粒度的列车运行数据和机器学习技术，可以准确地计算现有时刻表的实际性能。这将使拥有准确信息的规划者能够根据真实世界的证据更快、更好地做出规划决策。我们的时间表分析工具将自动向规划者提供与规划过程高度一致的更新的见解和建议。该工具利用轨道上(轨道电路)和车队上(GPS和OTMR)数据，将为Network Rail和TOC团队提供一个综合视图。从根本上说，该工具将导致时间表规划的速度和质量发生阶段性变化，从使用有限的模拟和轶事经验转向基于充分证据的方法。