大数据驱动的高速铁路高可用性研究

As one of the name cards of China's "Going Out" strategy, the high-speed rail (HSR) has undergone rapid development in recent years. The operating speed and other technical indicators have been continuously improved, meanwhile this renders the availability of HSR increasingly important. HSR availability is directly related to its reliability and maintainability, and has a significant impact on its safety, economy and sustainability. Chinese HSR has several years of operating experience and accumulated large amount of first-hand data with wide coverage, which serves a good foundation for big-data-driven theoretical research on its availability. On the other hand, the complex and dynamic natures of HSR big-data pose challenges to the new theoretical studies. This project will stand on the cutting-edge of availability theories, make use of HSR big-data, focus on the latest problems and challenges raised from HSR system, and then conduct theoretical innovation and empirical investigation. In particular, this project will establish a methodological framework covering the complete life cycle of HSR system from the following four aspects: system reliability design optimization and upgrading, system status online monitoring and fault pre-warning, system reliability evaluation and prediction and HRS maintenance strategy implementation. Our team members have accumulated considerably sufficient theoretical results and practical experiences in the above mentioned aspects. In addition, we have started collaborating with the research teams in Europe and other developed countries/regions. Therefore our team possess the capability to very effectively carry out this project.

作为我国“走出去”战略的一张名片，高铁近年取得了长足发展。在高铁运行速度等技术指标取得不断进步的同时，对高铁可用性的理论研究也变得日趋重要。高铁的可用性与可靠性、维修性直接相关，并对其安全性、经济性和可持续性发展有重大影响。我国高铁已有多年运营经验，积累的一手数据数量大且覆盖广，这为大数据驱动的高铁可用性理论研究提供了良好基础。另一方面，高铁大数据的复杂性、动态性等特点也对新的可用性理论研究提出了挑战。本课题将立足于可用性相关理论最前沿，基于高铁大数据，专注于高铁系统的新问题和挑战，开展理论创新和实证研究。具体而言，本课题将从系统可靠性优化设计与升级、系统状态在线监控与故障预警、系统可靠性评价与预测、高速铁路维修策略实现四个方面建立起覆盖高铁全生命周期的方法论体系。本团队成员在上述方面已经积累了相当充分的理论研究成果和应用经验，和欧美相关团队已开展合作，能非常有效的完成本课题的研究内容。

中国高铁经过近些年的快速发展已经成为世界上规模最大，运营里程最长，服务人数最多的交通基础设施，于此同时也对高铁系统长期高效、经济、稳定和安全运营提出了新的挑战。本项目以高铁系统高可用性为研究目标，对高铁系统从设计、监控、可靠性评价、故障预测和维修等开展覆盖其全生命周期的研究。.主要研究内容包括：（1）高铁系统可靠性优化设计与升级；（2）高铁系统状态在线监控与故障预警；（3）高铁系统可靠性评价与预测；（4）高速铁路维修策略。.项目团队目前已经在多个研究方法取得了丰富的理论成果。具体包括：基于多状态、寿命与参数不确定性的多目标串并联系统冗余分配问题；基于函数型数据的在线监视和系统潜在变化识别算法设计；非平衡样本下、未知和噪声标签下的系统故障检测问题；基于分布外数据检测的可信故障诊断，基于深度域适应的未知工况下故障诊断；高速列车关键组件可靠性评价；系统可靠性建模与评价；备件更换与维修策略；基于多目标优化的维修策略，系统故障后的韧性优化与恢复策略等。.部分研究成果已在有关单位应用落地，包括：高铁列车空心轴探伤周期优化和高铁列车车轮多边形识别与维修，已取得较高的经济效益。本项目的理论研究覆盖了高铁系统全生命周期，丰富了系统可用性理论，对中国高铁安全可靠运营已发挥积极作用。