Abstract
Empirical and mechanistical modeling approaches are often used in order to analyze functional relationships between process factors and system response and to identify process optima. The Kriging method allows to integrate both modeling approaches by combining statistical information on a given data set with a priori defined trend functions. However, trend functions from biotechnology applications are typically nonlinear with respect to the model parameters, which is not supported by standard Kriging. In this paper, we present an extension of the Kriging method for handling nonlinear trend functions by a Taylor based linearization approach which leads to an iterative parameter estimation procedure.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call