Page 266 - Distributed model predictive control for plant-wide systems

240 Distributed Model Predictive Control for Plant-Wide Systems

relatively fast systems, centralized MPC has been gradually replaced by decentralized or dis-
tributed MPC. DMPC accounts for the interactions among subsystems. Each subsystem-based
MPC in DMPC, in addition to determining the optimal current response, also generates a pre-
diction of future subsystem behavior. By suitably leveraging this prediction of future subsys-
tem behavior, the various subsystem-based MPCs can be integrated and therefore the overall
system performance is improved. Thus the DMPC is a good method to control HSLC.
Consider that the HSLC process is a large-scale system and each subsystem is coupled with
its neighbors by states, the DMPC framework should be suitable for the nonlinear system with
fast computational speed, appropriate communication burden, and good global performance.
Among the DMPC formulations provided in the literatures [27, 29, 33, 38, 41, 47, 122–125],
to guarantee performance improvement and the appropriate communication burden among
subsystems, the impacted region optimization-based DMPC is adopted here, where the opti-
mization objective of each subsystem-based MPC considers not only the performance of the
local subsystem but also those of its neighbors.
In this chapter, each subsystem-based MPC of the DMPC framework proposed is formulated
based on the successive online linearization of nonlinear model to overcome the computational
obstacle caused by nonlinear model. The prediction model of each MPC is linearized around
the current operating point at each time instant. Neighborhood optimization is adopted in each
local MPC to improve the global performance of HSLC and lessen the communication burden.
Furthermore, since the strip temperature can only be measured at a few positions due to the
tough ambient conditions, extended Kalman filter (EKF) is employed to estimate the transient
temperature of strip in the water-cooling section.
The contents are organized as follows: Section 11.2 describes the HSLC process and the
control problem. Section 11.3 presents the proposed control strategy of HSLC, which includes
the modeling of subsystems, the designing of EKF, the functions of predictor, and the devel-
opment of local MPCs based on the neighborhood optimization for subsystems, as well as the
iterative algorithm for solving the proposed DMPC. Both simulation and experiment results
are presented in Section 11.4. Finally, a brief conclusion is drawn to summarize the study and
potential expansions are explained.

11.2 Laminar Cooling of Hot-rolled Strip

11.2.1 Description

The HSLC process is illustrated in Figure 11.1. Strips enter cooling section at finishing
◦
rolling temperature (FT) of 820–920 C, and are coiled by coiler at coiling temperature (CT)
◦
of 400–680 C after being cooled in the water-cooling section. The X-ray gauge is used to
measure the gauge of strip. Speed tachometers for measuring coiling speed is mounted on
the motors of the rollers and the mandrel of the coiler. Two pyrometers are located at the
exit of the finishing mill and before the pinch roll, respectively. Strips are 6.30–13.20 mm
in thickness and 200–1100 m in length. The runout table has 90 top headers and 90 bottom
headers. The top headers are of U-type for laminar cooling and the bottom headers are of
straight type for low-pressure spray. These headers are divided into 12 groups. The first nine
groups are for the main cooling section and the last three groups are for the fine cooling