Page 27 - Distributed model predictive control for plant-wide systems
P. 27

1






             Introduction







             1.1  Plant-Wide System
             There is a class of systems which are composed of many interacted subsystems’ industrial
             fields. Especially with the development of the advanced technology and the increase in the
             requirement of products, many new distributed processes have appeared, the processes of pro-
             ducing products have become more and more complex, and the scales of industrial processes
             have become more and more large. The automation structure for this kind of systems has
             changed from the traditional centralized automation system to a decentralized and centralized
             automation system, and then to a distributed automation system.
               Correspondingly, the control algorithm and control structure for this kind of system change
             from centralized control and decentralized control to the distributed control system. The dis-
             tributed control refers to a control system where each subsystem is controlled by an individual
             controller, and these controllers communicate with other subsystem-based controllers and are
             coordinated according to the exchanged information for obtaining good global performance
             or some special common goals. So far, the distributed control, especially the DMPC, has been
             studied and are still being studied by many scientists, and many theories and algorithms have
             been developed. We think it is the right time to introduce the distributed control to more stu-
             dents and engineers.
               To make it more clear which kind of system is suitable for distributed control, we give some
             examples as follows.

             1. Wind power generation farm
               In a wind turbine power generation farm, as shown in Figure 1.1, wind turbines are spa-
               tially distributed. The output wind flow rate of each wind turbine decreases with increasing
               generated power. It affects the input wind flow rate of the downstream wind turbines, and
               then their dynamics. In this way, these wind turbines interact with each other. For the
               automation system, each wind turbine is controlled by an individual controller. And these
               controllers are connected by a network (fieldbus) and are able to communicate with each
               other by the network.






             Distributed Model Predictive Control for Plant-Wide Systems, First Edition. Shaoyuan Li and Yi Zheng.
             © 2015 John Wiley & Sons (Asia) Pte Ltd. Published 2015 by John Wiley & Sons (Asia) Pte Ltd.
   22   23   24   25   26   27   28   29   30   31   32