Intelligent alarm monitoring and management of complex industrial facilities

该联盟资助项目旨在开发、评估和应用先进的解决方案和工具，以帮助工业工厂运营商在异常情况下做出决策，从而提高工厂的安全运营。在工业设施运营中，制定工厂运营战略以实现可持续的环境、健康和安全（EH&S）绩效和企业社会责任变得越来越重要。这些EH&S目标还可以带来巨大的经济效益，因为由于计划外停机、设备损坏和操作故障，异常情况每年在行业中造成数十亿美元的损失。为了确保工业工厂的良好运行，并满足质量、效率和EH&S性能的预期措施，通过报警系统及时通知操作员工厂中的任何故障和异常情况，以便采取适当的补救措施，这一点至关重要。这种操作员决策支持系统具有故障检测与诊断和报警管理与监控两大功能，迫切需要先进的、智能的、实用的、可实施的解决方案。采用定量数据驱动的方法，以历史过程数据、报警和事件日志为基础，该提案概述了许多新的研究领域，用于复杂工业系统的先进和智能解决方案和工具开发，包括故障检测和诊断，在线报警洪水监测，操作员行为分析、多变量报警系统和早期根本原因诊断。新成果计划在参与伙伴的设施上进行工业验证和实施。将开发的解决方案和工具有可能改善合作伙伴以及加拿大各工业部门的运营、安全、生产和成本效益。

This alliance grant project is to develop, evaluate, and apply advanced solutions and tools for assisting industrial plant operators in their decisions on abnormal situations, and therefore enhance safe plant operations.In operating industrial facilities, it is becoming increasingly important to develop plant operation strategies for sustainable Environmental, Health, and Safety (EH&S) performance and corporate social responsibility. Such EH&S targets can also lead to enormous economic benefits since abnormal situations cost billions of dollars annually in industry due to unplanned shutdowns, equipment damages, and operation failures. To ensure satisfactory operation of an industrial plant and to meet desired measures of quality, efficiency and EH&S performance, it is of paramount importance that operators be informed promptly via the alarm system of any faults and abnormal situations in the plant so that appropriate remedies can be applied. Such an operator decision support system has two main functions, fault detection and diagnosis, and alarm management and monitoring, for which there exist strong needs for advanced and intelligent solutions which are practical and implementable.Adopting a quantitative data-driven approach, based on historical process data, and alarm and event logs, this proposal outlines a number of new research areas for advanced and intelligent solutions and tool development for complex industrial systems, ranging from fault detection and diagnosis, online alarm flood monitoring, operator action analysis, multivariable alarm systems, and early root cause diagnosis. The new results are planned to be industrially validated and implemented on facilities of participating partners. The solutions and tools to be developed have the potential for improving operations, safety, production and cost performance for the partners, as well as various Canadian industrial sectors in general.