To address the lack of rapid and flexible peak regulation capability in supercritical thermal power units, and the limitations of existing techniques (e.g., PID-based AGC with large lag, single-model MPC with poor adaptability to coal variations), this study proposes an Industrial Internet of Things (IIoT)-enabled intelligent control system. At its core, this system integrates an advanced multi-model predictive control algorithm, designed for processes with significant time delays and enhanced adaptability. Crucially, the system is built upon a wireless sensor network infrastructure that provides real-time, flexible data acquisition from key unit components. This architecture enables a novel multi-model predictive control strategy for the unit-boiler coordination system during peak regulation, significantly improving coordination and adaptive control performance. Experimental results under complex operating conditions—including rapid load changes, variable coal types, and high proportions of economical coal blending—demonstrate that this IIoT-enabled system effectively balances the speed and stability of coordinated control. Compared with the original PID-based CCS, the system increases the AGC comprehensive performance indicator ( K p ) from 2.0 to 4.5, enables stable operation at 30% rated load (180 MW) during deep peak regulation, reduces load response lag time by 35%, and controls main steam pressure/temperature dynamic deviations within ± 0.5 MPa and ± 4 °C, respectively. The unit’s comprehensive frequency regulation performance meets all target requirements. This work confirms that the integration of wireless sensing, IIoT architecture, and intelligent predictive control can substantially enhance the dynamic response and operational stability of thermal power units, offering a practical pathway for implementing Industry 4.0 solutions in power generation and broader industrial automation.
Zhang et al. (Thu,) studied this question.