摘要：AbstractThis paper presents modeling and parameter estimation of a thermal system, which is often nonlinear, with nonlinearities found in its parameters. As a result, it requires an additional online parameter estimation. In this paper, we implement an Extended Kalman Filter for modeling and parameter identification of such a device. The thermal device that we use is an educational device developed in-house. First, we derive the device’s mathematical model using a zero-order energy balance model as the template model. Next, we find the model’s best parameters by performing an optimization process. Finally, we implement an Extended Kalman Filter technique to accommodate the possibility that those parameters may change over time. Our experiments show that when we have reliable measurements and a reliable system model, the Kalman filtering technique can perform well as an online parameter estimator and act as a basis to build an adaptive model.
关键词：KeywordsRecursive identificationprocess modelingidentificationembedded computer control systemsapplications
