【作者】 王庆锋；

【导师】 高金吉；

【作者基本信息】 北京化工大学 ， 化工过程机械， 2011， 博士

【摘要】 中国过程工业企业设备管理基本属于传统的事后维修模式,设备基础管理薄弱,依靠经验的、定性的方法确定设备检查/维护内容,缺乏对关键设备的识别和分类,维修资源不能合理分配,存在“维修不足”和“维修过剩”,系统安全生产隐患大、事故多,设备可靠性、可用性和安全性难以控制和保证。为保证过程工业安全生产减少安全事故和环境事故,建立了过程工业基于风险和状态的设备智能维修决策及任务优化系统,它以基于风险和状态的设备完整性管理系统为架构,利用物联网技术和面向服务的架构技术综合集成了企业资源计划系统(Enterprise Resource Planning, ERP)、企业生产执行系统(Manufacturing Executive System, MES)和基于风险的维修(Risk Based Maintenance, RBM)系统与状态监测预知维修信息系统(Predictive Maintenance Information Systems, PMIS),为企业提供设备动态风险等级数据、预知维修信息数据、设备绩效指标数据、可靠性预测数据、设备剩余工作寿命数据,使各级人员能够通过网络平台及时准确地掌握设备风险状态和优化的维修任务排程,为设备维修决策提供科学支持。主要研究工作概括如下：(1)基于风险的维修研究和软件开发根据石油炼制、石油化工等过程工业设备管理特点,研究了适合过程工业设备管理特点的基于风险的维修风险评价技术,开发了基于风险的维修(RBM)软件,建立了基于风险的维修决策模型。(2)基于风险和状态的智能维修决策系统优化研究利用物联网技术搭建了基于风险和状态的设备智能维修决策系统。利用计算机技术、服务接口技术、数据库技术、有线或无线网络技术,基于面向服务的架构(Service Oriented Architecture, SOA)综合集成了PMIS、MES、RBM等系统模块,搭建了“以风险管理和核心,以专业管理为主线”的设备智能维修决策平台。该系统能够提供设备预知维修决策指标、动态风险等级指标、关键绩效指标,为基于风险和状态的维修决策提供定量分析数据。(3)基于风险的维修动态评估和设备管理绩效指标研究针对具体设备类型研究了可靠性数据、维修数据采集和交换内容；针对过程工业设备管理特点研究了设备动态风险变化影响因子(管理因子、个别设备修正因子)和动态风险评价技术；在对设备故障数据、维修数据进行分析的基础上,研究了过程工业设备绩效指标评估体系(设备、装置和公司三级)和设备管理绩效指标决策模型、绩效指标可靠性预测模型。(4)基于风险和状态的智能维修任务优化研究利用威布尔分布分析工具对平均故障间隔时间(Mean Time Between Failure, MTBF)、可靠性(Reliability)等动态数据进行分析,实现了设备的可靠性预测；利用主元分析方法确定设备故障特征参数,基于人工神经网络、灰色理论、曲线回归拟合和时间序列建模等方法跟踪设备故障特征参数劣化趋势,实现了设备剩余工作寿命预测；利用可靠性预测和剩余工作寿命预测实现了设备维修内容和故障维修间隔周期的优化；在设备动态风险分析和预知维修决策指标模型、管理绩效决策指标模型、设备可靠性预测模型、剩余工作寿命预测模型建立的基础上,以基于风险和状态的智能维修决策系统为平台,建立了基于风险和状态的设备维修任务优化模型。(5)基于风险和状态的设备智能维修决策系统工程应用基于风险和状态的设备智能维修决策系统综合集成了企业现有的ERP、MES、PMIS等设备管理信息资源,既考虑企业传统的设备管理现状,又引进了先进的RBM等设备风险管理技术；既考虑到设备定量风险分析缺乏可靠性数据和维修数据,又考虑到建立设备管理绩效指标的重要性；既强调建立的系统要具有动态风险等级指标、预知维修指标、设备管理绩效指标、可靠性和剩余工作寿命预测指标决策模型实现智能维修决策,又强调领导的强力支持、持续的培训和教育是基于风险和状态的维修管理模式成功应用的保证。锦州石化公司工程应用实践表明：建立的设备维修智能决策信息系统对于提高设备可靠性、可用性和安全性产生了积极效果,它使设备故障频率降低、故障后果减小,维修资源得到了合理利用。更多还原

【Abstract】 Equipment management in Chinese process industry mostly belongs to the traditional breakdown maintenance pattern, and the basic inspection/maintenance decision-making is insufficient. Equipment inspection/maintenance tasks are mainly based on empirical or qualitative methods, which usually lack identification and classification of critical equipment, so that maintenance resources can’t be reasonably allocated. Reliability, availability and safety of equipment are difficult to control and guarantee due to the existing maintenance deficiencies, maintenance surplus, potential danger and possible accidents. In order to ensure stable manufacturing and reduce operation cost, a risk & condition based equipment intelligent maintenance and decision-making optimization system is established in this paper, which utilizes risk & condition based equipment integrity management system as the architecture, and integrates ERP, MES (Manufacturing Executive System), RBM (Risk Based Maintenance)and PMIS (Predictive Maintenance Information System) through IOT (Internet Of Things) and SOA. This system can provide dynamic risk rank data, predictive maintenance information data, equipment performance index data, reliability prediction data and equipment residual life data, thus making personnel at all levels master equipment risk rank and optimized maintenance tasks in time and providing scientific support to maintenance decision-making. The main contents in the paper include:(1) Research on risk-based maintenance and software developmentAccording to the characteristics of equipment management in process industry such as petroleum refining and petrochemical enterprises, this paper investigates relevant risk-based maintenance risk evaluation methods, develops risk-based maintenance (RBM) software and establishes risk-based maintenance and decision-making model.(2) Research on risk & condition based intelligent maintenance and decision-making system optimizationUsing IOT, a risk & condition based intelligent maintenance and decision-making system is set up. Meanwhile, PMIS, MES and RBM modules are integrated on the basis of SOA by adopting computer technology, interface technology, database technology and cable/wireless technology, so that the equipment intelligent maintenance and decision-making platform (characterized by "risk management as the core, professional management as the main line") is formulated. This platform can provide predictive maintenance and decision-making indicator, dynamic risk rank indicator, key performance indicator as well as quantitative analysis data.(3) Research on risk-based dynamic maintenance evaluation and equipment management performance indexFor specific equipment types, data collection and data exchange of reliability data and maintenance data are researched; for the management characteristics in process industry equipment, dynamic risk variation influence factors (including management factor and individual equipment modifying factor) and dynamic risk evaluation techniques are investigated; on the basis of failure data and maintenance data, equipment performance indicator evaluation system (comprising equipment level, device level and company level), equipment management performance indicator decision-making model and performance indicator reliability prediction model are all studied.(4) Research on risk & condition based intelligent maintenance task optimizationUsing Weibull Distribution probability analyzing tool, analysis of MTBF (mean time between failure) and reliability data are carried out, thus realizing reliability prediction; using PCA (Principal Component Analysis), equipment failure features are determined; moreover, we trace the degradation trend of failure features by neural network, grey theory, curvilinear regression and time series modeling, thus realizing equipment residual life prediction; equipment maintenance content and maintenance period are optimized thanks to reliability prediction and residual life prediction; based on dynamic risk analysis and predictive maintenance and decision-making indicator model, management performance decision-making indicator model, equipment reliability prediction model and residual life prediction model, with risk & condition based intelligent maintenance and decision-making system as the platform, a risk & condition based intelligent maintenance task optimization system is established.(5) Engineering application of risk & condition based intelligent maintenance and decision-making systemThe system combines the existing ERP, MES, EAM and PMIS resources, and pays special attention to many aspects. Firstly, it considers traditional equipment management status, and introduces advanced RBM, RBI and SIL risk management techniques. Secondly, it takes into account the lack of reliability data and maintenance data in quantitative risk analysis, and gives attention to the importance of setting up management performance indicator. Thirdly, it emphasizes that the system should have the ability to realize intelligent decision-making based on dynamic risk rank indicator, preventive maintenance indicator, management performance indicator and residual life prediction indicator model, while advocating firm support from the leaders and continual training and education, which are important to the successful application of the system. The engineering application case in Petrochina Jinzhou Petrochemical shows that the establishment of the risk & condition based intelligent maintenance and decision-making system has brought about positive effect on the reliability, availability and safety of the equipment, lowering failure frequency, minimizing failure consequence and reasonably allocating maintenance resources.更多还原