Data-based Control of Process Feature Dynamics through Latent Behaviours

通过潜在行为对过程特征动态进行基于数据的控制

• 批准号: DP220100355
• 负责人: Prof Jie Bao
• 金额: $ 29.13万
• 依托单位: The University of New South Wales
• 依托单位国家: 澳大利亚
• 项目类别: Discovery Projects
• 财政年份: 2022
• 资助国家: 澳大利亚
• 起止时间: 2022-01-01 至 2024-12-31
• 项目状态: 已结题
• 来源: https://dataportal.arc.gov.au/RGS/Web/Grants/DP220100355
• 关键词: Data; based; Control; Process; Feature

This project aims to develop a novel data-based approach to control the feature dynamics of complex industrial processes. The dynamic features of desired process operations (leading to high energy and material efficiencies and good product quality) are often not directly measured but can be distilled from high-dimensional big process data. However, little effort has been made to develop process control approaches to achieve desired dynamic features. This project aims to develop such a data-based approach by controlling latent variable dynamics, using the behavioural systems framework integrated with big data analytics and artificial neural networks. The outcomes are expected to help build a cornerstone for future smart manufacturing.

该项目旨在开发一种新颖的基于数据的方法来控制复杂工业过程的特征动态。所需过程操作的动态特征（导致高能源和材料效率以及良好的产品质量）通常无法直接测量，但可以从高维大过程数据中提取。然而，很少有人致力于开发过程控制方法来实现所需的动态特性。该项目旨在通过控制潜在变量动态、使用与大数据分析和人工神经网络集成的行为系统框架来开发这种基于数据的方法。研究成果有望为未来智能制造奠定基石。