Real-time global optimisation for distributed parameter control systems

分布式参数控制系统的实时全局优化

• 批准号: DP190102478
• 负责人: Prof Ryan Loxton
• 金额: $ 26.69万
• 依托单位: Curtin University
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Projects
• 财政年份: 2021
• 资助国家: 澳大利亚
• 起止时间: 2021-01-01 至 2024-12-31
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DP190102478
• 关键词: Real; time; global; optimisation; distributed

This project aims to develop real-time optimal control algorithms for distributed parameter systems involving both time and spatial variables and multiple time-delays, with a focus on mining and energy applications. Current optimal control algorithms for such systems are too slow for real-time use and often get trapped at local optima, which can be vastly inferior to the global solution. This project will result in a new optimal control framework, underpinned by recent advances in constraint propagation, switching surface optimisation, and input regularisation. It will result in cutting-edge mathematical tools to complement and exploit new technologies and optimise key processes in natural gas liquefaction and zinc and alumina production, increasing efficiency and reducing the ecological footprint. This project will lead to new cutting-edge control algorithms for replacing the inefficient manual operations endemic in Australia’s natural gas and mineral processing plants.

该项目旨在为涉及时间和空间变量以及多时滞的分布参数系统开发实时最优控制算法，重点是采矿和能源应用。目前用于此类系统的最优控制算法对于实时使用来说太慢，并且经常陷入局部最优，这可能大大劣于全局解。该项目将产生一个新的最优控制框架，以约束传播，切换表面优化和输入正则化的最新进展为基础。它将产生尖端的数学工具，以补充和利用新技术，优化天然气液化和锌和氧化铝生产的关键工艺，提高效率并减少生态足迹。该项目将带来新的尖端控制算法，以取代澳大利亚天然气和矿物加工厂中普遍存在的低效手动操作。