Multirate Networked Control of Plantwide Chemical Processes

Multiple sampling of processes is a common practice in most large scale chemical industry. This together with the presence of interactions has often made controlling large scale chemical plants expensive and difficult. This paper aims to address these issues by assigning a common sampling window to account for the multi-sampling rates of processes while using a network approach to model large scale plant as a network of interconnected process units. Subsequently, the network is controlled by another network made up of interconnected local controllers. Using the concept of dissipativity, plantwide stability and global performance is then translated in conditions for which each controllers at different sampling instants has to satisfy. The controllers are then designed individually to allow a more autonomous and distributed control system.