Application of Varying Parameters Modelling with Gaussian Processes

AbstractThis paper describes an application of the method for modelling nonlinear dynamic systems from measurement data. The method merges the linear local model blending approach in the velocity-based linearisation form with Bayesian Gaussian process modelling. The new Fixed-Structure Gaussian Process model has a predetermined linear model structure with varying and probabilistic parameters represented by Gaussian process models. These models have several advantages for the modelling of local model parameters as they give us adequate results, even with small data sets. Furthermore, they provide a measure of confidence in the prediction of the varying parameters and information about the dependence of the parameters on individual inputs. The Fixed-Structure Gaussian Process model can be used for the extended local linear equivalence class of nonlinear systems. The modelling method is applied to modelling of a semi-industrial gas-liquid plant that exhibits variable dynamics depending on its operating region.