正交检测方法在钢铁铸造中的应用

The Application of Orthogonal Detection Method to Steel Casting

在设计一种基于预测控制器模型时，最关键的工作是识别一个合理可靠的过程模型．在钢铁厂的连续铸造线上识别铸模液位模型是很困难的，由于极高的随机干扰水平，使得众所周知的随机阶跃函数，ＰＲＢＳ或随机试验输入等参数识别方法都以失败告终．

The most critical task in designing a model based on predictive controller is identifying a reasonably good process model.The identification model of mould level of a continuous casting line in a steel plant is very difficult.Due to the extremely high stochastic disturbance level, well-known parametric identification methods using random step function, PRBS or stochastic test inputs are not successful.The present method based on orthogonal test inputs gives satisfactory results.