Process data analytics

目标 融合来自不同数据源的信息是制定智能分析平台基础策略的关键步骤。在分析在过程工业中的应用背景下，这个赠款提案是开发一个理论框架和工具，用于无缝集成过程和警报数据库中的信息，并补充过程连接信息。 从这些不同的和复杂的数据源中发现的信息可以随后用于过程和性能监控，包括报警合理化、过程故障的根本原因诊断、HAZOP分析、安全和最佳过程操作。这种多变量过程数据分析涉及从常规过程数据中提取信息，该常规过程数据通常是非分类的（如在来自传感器的数字过程数据中），加范畴的来自警报和事件（A&E）的（或非数值或定性和二进制）数据与过程连通性或拓扑结构信息相结合的日志，这些信息可以通过因果关系分析从数据中推断出来，或者从过程后者是指捕获过程单元中的物质流以及由于控制回路而导致的过程中的信息流路径。 新奇 高度互连的过程工厂现在很常见，并且对过程异常的根本原因的分析（包括预测性风险分析）是重要的。该提案的主旨是开发一个理论框架，用于使用统计推断方案从存档的过程数据中提取信息和知识，并将这些模型与警报数据和过程连接信息集成和验证。该提案的独特之处在于包含了基于数据的过程连接信息，用于过程监控，因此代表了过程数据分析的重大范式转变。这样的方法将作为一种使能工具，用于预测性和主动的过程资产维护以及安全和最佳的过程操作。 预期意义 目前，分析在不同领域（例如工程，医学等）的应用呈爆炸式增长。同样，流程工业目前存档的数据量非常大（大数据），本提案的主要目的是在这些数据中找到价值，并在线使用这些数据进行安全和最佳的流程操作。 该提案的社会经济意义将是积极的，而不是被动的，管理与高生产力和能源效率的工厂的过程操作相结合，这些工厂点缀着加拿大的景观，包括纸浆和造纸，石油化工，食品加工，发电，矿物加工和石油天然气勘探。该项目的一个同样重要的方面是教育和培训具有统计数据挖掘技能的人力，这是非常需要的。

Objectives The fusion of information from disparate sources of data is the key step in devising strategies for the foundation of a smart analytics platform. In the context of application of analytics in the process industry, this grant proposal is to develop a theoretical framework and a tool for seamless integration of information from process and alarm databases complimented with process connectivity information. The discovery of information from such diverse and complex data sources can be subsequently used for process and performance monitoring including alarm rationalization, root cause diagnosis of process faults, HAZard and OPerability (HAZOP) analysis, safe and optimal process operation. Such multivariate process data analytics involves information extraction from routine process data, that is typically non-categorical (as in numerical process data from sensors), plus categorical (or non-numerical or qualitative and binary) data from Alarm and Event (A&E) logs combined with process connectivity or topology information that can be inferred from the data through causality analysis or as obtained from piping and instrument diagrams of a process. The later refers to the capture of material flow streams in process units as well information flow-paths in the process due to control loops. Novelty Highly interconnected process plants are now common and the analysis of root causes of process abnormality including predictive risk analysis is non-trivial. The thrust of this proposal is to develop a theoretical framework for extracting information and knowledge from archived process data using statistical inference schemes and integrating and validating such models with alarm data and process connectivity information. The unique aspect of this proposal is the inclusion of data-based process connectivity information for process monitoring and thus represents a major paradigm shift in process data analytics. Such a methodology would serve as an enabling tool for predictive and pro-active process asset maintenance and safe and optimal process operation. Expected significance There is currently an explosion of applications of analytics in diverse areas (e.g. engineering, medicine, etc). In the same vein the volume of data currently archived by the process industry is massive (BIG Data) and the key aim of this proposal is to find value in this data and use this on-line for safe and optimal process operation. The socio-economic significance of this proposal will be pro-active, as opposed to reactive, management combined with highly productive and energy efficient process operation of plants that dot the Canadian landscape including pulp&paper, petro-chemical, food processing , power generation, mineral processing and oil and gas exploration. An equally important aspect of this project is the education and training of manpower with statistical data mining skills that are in high demand.