Decoupling Predictive Control of Reheat Steam Temperature of USC Double Reheat Unit

Based on the actual operating results of the domestic and foreign USC(ultra-supercritical) double reheat units, the USC double reheat unit has better economic and environmental benefits than one reheat unit when they have same capacity. However, the reheat steam temperature of double reheat unit becomes a multivariable system which is characterized by strong coupling, big inertia, and time delay, due to the complexity and particularity of double reheat unit's boiler construction. Realizing the effective control of two reheat steam temperatures requires a prompt solution. Predictive control is an effective method against such characteristics as time delay and big inertia-controlled object. This paper presents a decoupling predictive control strategy with input constraints to control the reheat steam temperature. To measure the control quality of the reheat steam temperature control strategy, the increasing and decreasing load tests were analyzed on the emulation platform of 1000MW USC double reheat unit, using Xi'an Thermal Power Research Institute adopting APROS emulation software. Simulation results show that the decoupling predictive control strategy with input constraints can achieve a better performance (except disturbance rejection) when compared with PID (Proportion Integration Differentiation).