一种实用的磨矿多模型预测控制策略研究

Practical Multiple Model Strategy for Multivariable Model Predictive Control

磨矿过程是选矿过程中的重要一环,属多输入多输出系统,并存在严重的扰动变量(诸如硬度等)和时变参数。介绍了一种磨矿控制系统中实用的多变量动态矩阵控制的多模型控制策略。该方法基于不同的矿石硬度对球磨过程建立不同的阶跃模型进行多模型预测控制。仿真证实了该方法的有效性。

The grinding process is one of the key of the benefication process.The process is a Muti-input and Muti-output system with severe disturbances（such as mill hardness） and time-variable parameters.The objective of this work is to introduce a multiple model control strategy of multivafiable dynamic matrix control for ball mill grinding circuit.The method is to construct different by step models based on different hardness of the mill to complete the multiple model predictive control.