改进型带约束的预测控制算法在超超临界机组协调控制系统中的应用

Improved Predictive Control Algorithm with Constraints in Coordinated Control System of Ultra-supercritical Unit

基于某超超临界机组协调控制系统的被控对象,以及Pade近似的思想对被控对象的模型进行降阶处理,在不改变被控对象原有动态特性的前提下,降低了系统的复杂性。在此基础上设计带约束的多变量广义预测控制器,并进行仿真试验,结果显示,设计控制器的计算量得到降低,相对于传统PID算法,机组实发功率、主汽压力、中间点温度响应设定值的速率明显加快,超调量及波动较小,燃料量变化及时,汽机调门开度和给水流量在一定的约束条件下变化得到抑制,提高了机组的稳定性和经济性。

Based on the controlled object of an ultra-supercritical unit coordinated control system,this paper reduces the model of the controlled object model based on the idea of Pade approximation,and reduces the complexity of the system without changing the original dynamic characteristics of the controlled object.On this basis,a multivariable generalized predictive controller with constraints is designed and simulated.