一种改进的Smith预估补偿方法  

An Improved Smith Predictor Control Method

作　　者：汪秉文[1] 李国宽[2] 姜新 

机构地区：[1]华中理工大学控制科学与工程系 [2]华中理工大学图像识别与人工智能研究所

出　　处：《华中理工大学学报》1999年第12期33-35,共3页Journal of Huazhong University of Science and Technology

摘　　要：在常规Ｓｍ ｉｔｈ 预估补偿算法控制结构的基础上，进行适当改进，使用模型参考自适应辨识方法，把控制对象与模型间的差异用增益来补偿，使得模型失配条件下的Ｓｍ ｉｔｈ 预估补偿控制效果接近于模型精确条件下的控制效果．仿真结果表明，在存在模型偏差时，提出的方法能大大改善Ｓｍ ｉｔｈ 预估补偿控制的动态性能，且对外扰具有很好的抑制作用．Appropriate improvements on conventional Smith predictor control structure have been made. All discrepancies between control object and model are compensated by gain. By means of model reference adaptive identification method, the gain of the model is accordingly adjusted, so that the control quality in case of model miss matching is close to that of systems using the Smith predictor with accurate model. Simulation results indicate that the method presented can significantly improve the dynamic behavior of the control with Smith predictor, and efficiently suppress any external disturbance.

关 键 词：延迟系统 SMITH预估器 模型参考 自适应辨识 

分 类 号：TP273.2[自动化与计算机技术—检测技术与自动化装置]