【摘要】 面对工业系统日趋复杂且安全挑战严峻的紧迫局势，传统被动模式难以应对突发性系统级风险，向智能运维（AIOps）转型成为保障全局安全的必然诉求。为了破解多学科、多要素耦合的复杂难题，本文突出模型化的研究手段，基于系统工程视角构建了涵盖资源、组织、信息等六大维度的AIOps场景架构模型，深度剖析了状态感知（PHM）与决策执行（MRO）的闭环协同机制。研究强调，单一数据驱动方法在复杂工况下存在局限，应对极端工况仍需专家机理兜底；同时，智能化转型不止一种道路，企业应依托模型指导探索契合自身的差异化演进路径。更多还原

【Abstract】 Against the backdrop of increasingly complex industrial systems and mounting safety challenges, traditional passive modes are no longer adequate for coping with sudden system-level risks. Transitioning toward intelligent operations(AIOps) has become an essential requirement for ensuring global safety. To address the multi-disciplinary and multi-factor coupling challenges, this study highlights modelling-centred analytical methods. From a systems-engineering perspective, it develops an intelligent-operations scenario architecture encompassing six dimensions, including resources, organisations and information, and provides an in-depth analysis of the closed-loop collaborative mechanism between prognostics and health management(PHM) and decision execution(MRO). The study emphasises that a single data-driven approach has limitations under complex operating conditions, and expert mechanisms are still required to address extreme conditions. Moreover, intelligent transformation does not follow a single trajectory; enterprises should instead explore differentiated evolution paths aligned with their own contexts under the guidance of architectural models.更多还原