NONLINEAR SYSTEM IDENTIFICATION OF A CLOSED-LOOP LEAN COMBUSTION PROCESS

摘要

This paper deals with nonlinear system identification of a set of measurements collected on a lean premixed combustion process operating in a limit cycle. Due to the fact that the data are collected in closed-loop and the system has no external excitation, the identification task is particularly challenging. This work mainly focuses on the issue of the feasibility of the identification task. It is shown that, despite the paucity of information available, a gray-box non-linear model can be estimated quite accurately. Specifically, the model proposed provides an accurate explanation both of the limit-cycle fundamental oscillation and of a non-harmonic high-frequency signal affecting the pressure of the combustion chamber.