Options
Distributed model predictive control of a system with multi-rate and delayed measurements
Date Issued
01-01-2018
Author(s)
Ravi, Arvind
Indian Institute of Technology, Madras
Abstract
The objective of this work is to implement a sequential distributed model predictive control (MPC) for a process with infrequent and delayed primary measurements in presence of disturbances in the system. The overall system comprises of multiple units. A centralized state estimator based on Kalman filter is used to estimate states for the entire system. The estimator uses sampled state augmentation approach for fusing the delayed information from primary measurements with the frequent secondary measurements to obtain better estimates for output feedback control. Distributed MPC is developed that uses these estimates, as well as selective information transfer, based on process knowledge, between different units of the system. An example of a reactor separator system is considered to demonstrate the applicability of the control formulation. This system also has a plant-model mismatch of some parameter values and a disturbance model framework is incorporated.
Volume
44