Superheated steam temperature (SST) regulation in large coal-fired combined heat and power (CHP) units becomes increasingly challenging under deep peak-shaving. Wide-range load variations and multisource disturbances require strong disturbance rejection, whereas industrial distributed control systems typically retain PID-oriented implementations with limited computational capability. Motivated by these, this paper proposes a bilevel IMC-DDE-PID control framework for SST regulation. In the inner loop, a desired dynamic equation (DDE)-based tuning rule shapes the plant dynamics to follow a prescribed desired closed-loop model without requiring a precise process model and partially attenuates input-side disturbances. The residual disturbance is then handled by an outer internal model control (IMC) layer built upon the desired model as an embedded internal model. Comprehensive simulations are conducted on seven benchmark SISO plants and on identified SST models at multiple loads, including scenarios with large load variations and cross-load model variation, to evaluate tracking, disturbance rejection, and control variation. Finally, field tests on an in-service 660 MW CHP unit under deep peak-shaving validate the practicality of the proposed method, reducing the SST fluctuation range by 7 °C.
Zhao et al. (Mon,) studied this question.