循环流化床锅炉的模型预测控制

Model Predictive Control of Circulating Fluidized Bed Boiler

利用系统辨识技术建立锅炉的预测模型,对模型预测控制算法在多变量系统中的应用进行了讨论,用于控制循环流化床的蒸汽压力、蒸汽温度和炉床温度.在MATLAB/SIMULINK环境下对模型预测控制系统进行了仿真.研究结果表明,利用系统辨识技术建立的系统模型结构简单,可以在有限时域内实现系统输出的精确预测,适合在线控制应用;模型预测控制算法适用于循环流化床锅炉的控制,其仿真控制结果明显优于经典控制系统.

The literature survey of CFBB modeling was carried out as priori knowledge for system identification. A control and prediction aimed experimental model was obtained via system identification techniques. Model predictive control （MPC）, as a model reference control algorithm, is discussed thereafter as a multivariable controller for steam pressure, steam temperature and bed temperature. The MPC system was simulated in MATLAB/SIMULINK. It is concluded that system identification technique is a handy tool in obtaining the prediction model of CFBB with simple model structure, and that MPC algorithm is suitable for controlling CFBB. The simulation control performance shows satisfactory characteristics compared with the PID control result.