Rahul Bindlish

Model order selection for process identification applied to an industrial ethylene furnace

Abstract This paper presents a quantitative analysis of the model order selection problem, and its application for system identification of an ethylene furnace with open-loop and closed-loop industrial plant data. Empirical ARX models are used to describe the physical phenomena in the ethylene furnace. Appropriate model order selection is done based on the information content in the industrial data from the ethylene plant. Model order is chosen by using Akaikes information criterion (AIC), Rissanens minimum description length (MDL), and a criterion based on the unmodeled output variation (UOV). UOV results in a smaller order model that has well-defined parameters with tight confidence intervals as compared to AIC and MDL. Similar models are obtained using closed-loop and open-loop data from the industrial process when UOV is used because the models are well-determined.

Nonlinear model predictive control of an industrial polymerization process

Abstract Nonlinear model predictive control (NMPC) is used to maintain and control polymer quality at specified production rates because the polymer quality measures have strong interacting nonlinearities with different temperatures and feed rates. Polymer quality measures that are available from the laboratory infrequently are controlled in closed-loop using a NMPC to set the temperature profile of the reactors. NMPC results in better control of polymer quality measures at different production rates as compared to using the nonlinear process model with reaction kinetics to implement offline targets for reactor temperatures.

Autocovariance-based plant-model mismatch estimation for linear model predictive control ☆

Abstract In this paper, we present autocovariance-based estimation as a novel methodology for determining plant-model mismatch for multiple-input, multiple-output systems operating under model predictive control. Considering discrete-time, linear time invariant systems under reasonable assumptions, we derive explicit expressions of the autocovariances of the system inputs and outputs as functions of the plant-model mismatch. We then formulate the mismatch estimation problem as a global optimization aimed at minimizing the discrepancy between the theoretical autocovariance estimates and the corresponding values computed from historical closed-loop operating data. Practical considerations related to implementing these ideas are discussed, and the results are illustrated with a chemical process case study.

Parameter estimation in a dynamic model of a copolymerization process

This paper presents the development of a mathematical model for a copolymerization process. The process model consists of the material balances for the reactor and simplified dynamics for the downstream separator. The dynamic model of the reactor consists of material balances including expressions for reaction kinetics. The kinetic expressions are simplified by using the quasi-steady-state assumption for active polymer chains and the moments of chain length distribution for the active and dead polymer chains. The model parameters are estimated using the maximum likelihood method. This model is augmented with an output disturbance model to obtain a closer representation of the actual polymerization process.

Plant-model mismatch estimation in unconstrained state-space MPC

In this paper, we develop an autocovariance-based method for estimating plant-model mismatch in unconstrained model predictive control systems using discrete-time, linear time-invariant state space models. We rely on knowledge of the process noise model, together with other reasonable assumptions, to derive an explicit expression for the autocovariance matrix of the closed-loop outputs. Then, we prove that by minimizing the discrepancy, in a norm sense, between this theoretical value and the autocovariance matrix of the operating data, we find an asymptotically consistent estimate for the plant-model mismatch. We illustrate the use of this approach with a case study.

Scheduling, optimization and control of power for industrial cogeneration plants

Abstract Scheduling, optimization and control of power for three industrial cogeneration plants at one of Dow’s Louisiana site is presented in this paper. A first principle mathematical model that includes mass and energy balances for gas turbines, heat recovery units, steam turbines, pressure relief valves and steam headers is used to formulate multiple optimization problems to recommend the best strategy to trade power. The model has detailed operational information that includes equipment status and control curves for different operating scenarios. The scheduled power offer curve is obtained by solving multiple optimization problems using the validated process model along with operational and equipment limitations. Adjustment of power schedule offer is done in the real-time market thirty minutes prior to the hour and implementation of the dispatched power schedule is done using a model predictive controller.