Identification of a gas furnace process using WSDP model

This paper presents the identification of a gas furnace process using the so called wavelet based State Dependent Parameter (WSDP) models. Here, each of the process' dominant dynamics are characterized by a state dependent parameter in form of wavelet which is able to provide valuable insight, and effective means to measure the system's degree of nonlinearity. To illustrate the advantages of the proposed modeling approach, the identified model is benchmarked with a few welknown reported results, i.e. fuzzy logic, generalised kernel and linear ARX models.