Performance monitoring and root-cause diagnosis of industrial model predictive control systems for sustainable energy production

可持续能源生产工业模型预测控制系统的性能监测和根本原因诊断

• 批准号: 437721-2012
• 负责人: Yu, Jie
• 金额: $ 2.6万
• 依托单位: McMaster University
• 依托单位国家: 加拿大
• 项目类别: Collaborative Research and Development Grants
• 财政年份: 2012
• 资助国家: 加拿大
• 起止时间: 2012-01-01 至 2013-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=498446
• 关键词: Performance; monitoring; root; cause; diagnosis

Industries are facing significant challenges on sustainable profits, resources, energy and environment. As the most popular advanced control technology, model predictive control (MPC) has been widely applied to different industries including energy, chemical, materials, mining, power, pulp & paper, and automotive industries. Significant economic profits, energy savings and environmental pollution reductions have been achieved from industrial MPC technology. However, MPC performance often encounters significant degradation, which can cause serious safety incidents, profit losses, reduced energy efficiency and increased environment pollutions in industrial productions. Currently, there are no comprehensive and effective solutions to monitor and particularly diagnose industrial MPC performance. This research project is aimed to develop the novel state-of-the-art techniques for monitoring complex MPC systems and further diagnosing the root causes of performance degradation. With the accurate and reliable diagnosis findings, corrective actions can be automatically taken to optimize the process operation with the best profitability and energy efficiency as well as the lowest environmental emissions and carbon footprints. Moreover, the MPC monitoring and diagnosis system will be integrated into MPC life cycle to facilitate and improve MPC design, commissioning and maintenance in industrial processes. The developed technology will be demonstrated in energy production processes at Shell and further be extended to different kinds of industries as generic technical solutions. It is anticipated that the research inventions of this project will significantly benefit a wide range of industries in Canada with considerable economic, environmental and social values, which include higher productivity, boosted energy efficiency, mitigated carbon emissions and reduced safety incidents. In addition, this research will make substantial knowledge contributions to chemical engineering and particularly process control field with the new findings on monitoring and control theories as well as complex system sustainability.

工业在可持续利润、资源、能源和环境方面面临重大挑战。 模型预测控制（Model Predictive Control，MPC）作为当今最流行的先进控制技术，已广泛应用于能源、化工、材料、矿山、电力、造纸、汽车等行业。工业MPC技术已取得了显著的经济效益、节约能源和减少环境污染。然而，MPC的性能往往会出现严重的下降，这可能会导致严重的安全事故，利润损失，降低能源效率和增加环境污染的工业生产。目前，还没有全面有效的解决方案来监测和特别诊断工业MPC性能。该研究项目旨在开发新的最先进的技术，用于监测复杂的MPC系统，并进一步诊断性能下降的根本原因。凭借准确可靠的诊断结果，可以自动采取纠正措施，以优化流程操作，实现最佳盈利能力和能源效率，以及最低的环境排放和碳足迹。此外，MPC监测和诊断系统将被集成到MPC生命周期中，以促进和改善工业过程中MPC的设计、调试和维护。开发的技术将在壳牌的能源生产过程中进行演示，并进一步扩展到不同类型的行业作为通用技术解决方案。预计该项目的研究发明将使加拿大的众多行业受益匪浅，具有可观的经济、环境和社会价值，包括提高生产力、提高能源效率、减少碳排放和减少安全事故。此外，这项研究将为化学工程，特别是过程控制领域提供大量知识，并在监测和控制理论以及复杂系统可持续性方面取得新的发现。