迭代学习控制与软测量技术在打浆过程中的应用  

Application of Soft Measurement Module on Pulp Beating Degree and Iterative Learning Control in Pulp Beating Process

作　　者：马汇海[1] 张根宝[1] 李强华[1] 姜丽波[1] 

机构地区：[1]陕西科技大学电气与电子工程学院,陕西咸阳712081

出　　处：《计算机测量与控制》2006年第12期1620-1622,共3页Computer Measurement &Control

摘　　要：打浆过程机理复杂,影响因素众多,数学模型提取困难,导致控制系统的控制质量不易提高;将打浆度软测量模型和迭代学习控制算法相结合,并且适当利用成熟经验,采用比例积分型迭代学习规律,设计出可行的控制方案和算法流程,在迭代学习中不断校正学习参数,逐渐缩小软测量模型与真实过程间的误差,解决了由于模型不清晰导致的控制质量较差的问题;经仿真实验证明,该控制方案能有效提高控制系统的准确性和快速性,具有良好的控制效果。Because of the complexity in the process of pulp beating and numerous influence factors and the difficulty in the foundation of the mathematic model, it is hard to improve the control quality of control system. This paper offered a new way based on the soft measurement model of pulp beating degree and iterative learning control. With ripe experience appropriately and PI-type iterative learning rule, designed feasible control scheme and algorithm procedure. During the process of iterative learning, the learning parameter is adjusted in order to reduce the difference between the soft measurement model and the actual process and resolved the problem of inferior control quality caused by the blurry mathematic model. The simulation testing result has proved that the eontrol method can improve the aeeuraey and rapidity of the control system effectively and get good performanee.＂

关 键 词：迭代学习控制 软测量 打浆度 

分 类 号：TP273.22[自动化与计算机技术—检测技术与自动化装置]