Big Data-based Distributed Control using a Behavioural Systems Framework

随着工业4.0成为现实，工业过程正在成为生成、处理、存储和通信大量数据的分布式网络物理系统。利用行为系统框架，该项目旨在开发一种新的分布式控制方法，直接基于大数据的复杂过程。开发了一个新的无模型框架来表示和分析过程/控制器网络和相互作用的影响，并确定分布式控制下期望控制性能的可行性。通过扩展行为系统的耗散性、收缩和微分耗散性条件，将开发新的基于大数据的分布式控制设计方法。

With Industry 4.0 turning into reality, industrial processes are becoming distributed cyber-physical systems which generate, process, store and communicate large amounts of data. Using the behavioural systems framework, this project aims to develop a novel distributed control approach for complex processes directly based on big process data. A new model-free framework will be developed to represent and analyse the process/controller networks and interaction effects, and determine the feasibility of desired control performance under distributed control. Novel big data-based distributed control design approaches will be developed by extending the dissipativity, contraction and differential dissipativity conditions for behavioural systems.