MODEL IDENTIFICATION IN RAPID THERMAL-PROCESSING SYSTEMS

Of the various techniques for controlling the temperature in rapid the rmal processing (RTP), model-based control has the greatest potential for attaining the best performance, when the model is accurate. In thi s paper, some so-called system identification methods are introduced t o help obtain more accurate models from measured input-output data. In the first identification, techniques are presented to estimate the pa rameters (time constant and gain) of a particular physics-based model. In the other, we also show how to use the input-output measurements t o obtain a black-box model (autoregressive exogenous model) of the RTP system, which turns out to have better predictive capability. For eac h problem, a brief summary of the theoretical derivation of the identi fication technique and assumptions on which it is based is given and e xperimental results based on data collected from an RTP system are des cribed. The identified model is useful not only for control law design but also to study certain system characteristics. Studying the so-cal led DC response in this way led us to reconfigure the chamber geometry of the existing RTP system to more effectively redistribute the light energy from the lamps.