预测函数控制在球磨机磨矿回路中的应用

Application of Predictive Functional Control on Ball Mill Grinding Circuit

球磨机磨矿回路具有多变量强耦合、大延迟和非线性等特性,生产工况变化性大,常规采用多回路PID解耦控制,使回路之间的相互耦合作用减到最小,但这种控制不能及时且精确地控制磨矿过程。模型预测控制能够解决多变量系统之间的耦合作用,而预测函数控制是在模型预测控制原理的基础上提出的,并且将控制规律进行了结构化设计,使控制器在线计算量小,参数调整简单,跟踪快速,具有很强的抗扰动性和鲁棒性。本文以实验型球磨机磨矿回路为研究对象,分别设计多回路PID解耦控制和预测函数控制,并对两种控制的仿真效果进行对比。仿真结果表明,该控制策略具有较好的跟踪性能、抗干扰能力,可以达到较为理想的控制效果。

Ball mills grinding circuits are muhivariable,strongly coupled,long time delay and nonlinear process.The operating conditions often vary violently. Traditionally grinding circuits are controlled by decouped multi-loop PID controllers that minimize the effect of interaction among the control loops. The PID controllers generally become sluggish and a accurate control of the circuit is not possible. Model predictive controllers can handle such highly interacting muhivariable systems effectually due to its coordinated approach. Predictive functional control （PFC）is a branch of the MPC family. The main distinguishing feature of the PFC algorithm over other MPC algorithms is the construction of the manipulated variables. So the controller＇s on line-calculation work is being reduced ,parameter adjustment facilitated, tracing ability improved,featured by good disturbance resistance and robustness at the same time.Simulation of PFC studies on a laboratory ball mill grinding circuit are carried out compared with decoupled multi-loop PID controllers, Predictive functional controllers and their performances are compared. The simulation result shows that the control strategy can achieve better control performance of setpoint tracking and disturbance rejection.