针对机炉协调控制系统中负荷变动时传统PID控制效果不佳、机组运行经济效益降低甚至影响系统安全状况的问题,提出一种采用变时域监督预测的机炉协调控制方法(VDMC)。首先根据所分析的超临界直流炉的机理特性,建立随煤质波动的变参数模型;再结合实际工况确定的约束条件以及电网频率的扰动,采用带监督环节的多变量受约束预测控制算法来解决控制框架中分散和集中的平衡性;最后基于变时域的思想,改善了机炉协调控制的整体性能。实验结果表明,与传统PID控制方法和固定时域预测控制方法 相比,所提出的VDMC方法的相对误差均值降低了20%左右,有效提升了机炉协调控制品质,确保了系统安全稳定运行。
A varying horizon supervisory predictive control algorithm (VDMC) is proposed to deal with the problem that the traditional PID controller is insufficient in coordination control systems (CCS), leading to lower economic benefit of power units or even affecting system safety during load changing. Firstly, a time-varying parameter model with fluctuation of coal quality is established through the analysis of mechanism characteristics of supercritical boilers. Then, a multivariable constrained supervisory predictive control algorithm is then used to solve the balance problem of centralization and decentralization in the control frame by taking constraint conditions determined by the actual situation and the disturbance of power grid frequency into account. Finally, the overall performance of the coordination control for turbine-boiler system is improved based on the idea of varying horizon. Simulation results and comparisons with the traditional PID and the fixed horizon predictive control algorithm show that the mean value of relative errors of the proposed method is reduced by about 20 % during the load changing period. The method enhances the quality of turbine-boiler coordination control and makes the system more stable.