Model-Based Process Optimization in the Presence of Parameter Uncertainty

Abstract

Abstract In model-based process optimization one uses a mathematical model to optimize a certain criterion, for example the product yield of a chemical process. Models often contain parameters that have to be estimated from data. Typically, a point estimate (e.g. the least squares estimate) is used to fix the model for the optimization stage. However, parameter estimates are uncertain due to incomplete and noisy data. In this paper, we show how parameter uncertainty can be taken into account in process optimization. To quantify the uncertainty, we use Markov Chain Monte Carlo (MCMC) sampling, an emerging standard approach in Bayesian estimation. In the Bayesian approach, the solution to the parameter estimation problem is given as a distribution, and the optimization criteria are functions of that distribution. We study how to formulate and implement the optimization and show by numerical examples that parameter uncertainty can have a large effect in optimization results.