A Novel Global Model Calibration Scheme for Complex Process Systems Using an Adaptive Experimental Design

First principle modeling is a crucial step before implementing model-based control and optimization of large-scale complex process systems. Due to lack of prior knowledge of process devices and technologies, the analysis of first principle models first requires tackling the calibration problem of uncertain parameters, especially when the model parameters change with operating conditions of the system. To ensure the extrapolation accuracy of model over the global operating domain and meanwhile reduce the experimental cost of condition testing, this work presents a global model calibration scheme based on an adaptive experimental design. The experimental conditions are selected by criteria of sampling density, local nonlinearity and parameter uncertainty to perform parameter estimation. To prove the effectiveness of our scheme, the algorithm in this work is applied to the case of high-temperature gas-cooled reactor pebble-bed module nuclear power plant and compared with some conventional experimental designs.