Dual Rate Adaptive Control for an Industrial Heat Supply Process Using Signal Compensation Approach

The industrial heat supply process (HSP) is a highly nonlinear cascaded process which uses a steam valve opening as its control input, the steam flow-rate as its inner loop output and the supply water temperature as its outer loop output. The model parameters, such as steam density, specific heat of steam, dirtiness resistance and coefficient of heat supply are usually unknown and nonlinear functions. Moreover, these model parameters vary in line with steam pressure, ambient temperature and the residuals caused by the quality variations of the circulation water. In this paper, the unknown variations of dynamics are represented as the combination of known previous sampled unmodeled dynamics and its changing rate. The combination signals aiming at eliminating the effects of the previous sampled unmodeled dynamics and tracking error are constructed and added onto the control signal from feedback controller based on linear deterministic model, and then a lifting technology is adopted to propose a novel compensation signal based dual rate adaptive controller. This controller is applied successfully to an industrial heat supply process, where it has been shown that both the supply water temperature and the rate of the changes of the steam flow-rate can be controlled well inside their targeted ranges when the process is subjected to unknown variations of its dynamics.