Research and Application of Multiple Model Predictive Control in Ultra-supercritical Boiler-Turbine System - Intelligent Computing for Sustainable Energy and Environment - page 474

Fig. 3. piecewise step response models at different load point of Ultra-supercritical
unit
simulation environment. MV1, MV2, MV3 represents the governor valve open-
ing, the fuel flow and the water flow respectively. CV1, CV2, CV3 represents
the load, the steam temperature and the steam pressure respectively. Load is
constrained by setpoint control while steam temperature and steam pressure is
constrained by zone control. Figure 4 and Figure 5 showed the manipulated vari-
ables and controlled variables in simulation when load changed from 1000MW
to 900MW. In Figure 4, the red curve, orange curve and blue curve represented
the governor valve opening, the water flow, the fuel flow respectively. In Figure
5, the red curve, orange curve, blue curve represented the load, steam pressure
and steam temperature respectively. Figure 6 was the figure of change of load.
Figure 7 was the figure of change of the steam temperature. Figure 8 and Figure
9 showed manipulated variables and controlled variables respectively when the
load changed from 600MW to 700MW.
The results show that the value of CV could rapidly track the expected value
of CV in the process of changing load and the setpoint control was accurate in the
steady working condition. The load could be changed by the rate of 20-30MW
per minute. Steam pressure smoothly rose and dropped following load. Steam
temperature fluctuated in the range of 1 degrees around of setpoint 599. The
system could guarantee good static and dynamic performance. So the multiple
models MCPC with double-layer structure is a promising method for coordina-
tion control of Ultra-supercritical unit.