一种改进的模糊Smith预估器  

An Improved Fuzzy Smith-predictor

作　　者：王舰[1] 强文义[1] 麻亮[1] 张福恩[1] 

机构地区：[1]哈尔滨工业大学控制科学与工程系,黑龙江哈尔滨150001

出　　处：《控制工程》2005年第4期331-334,共4页Control Engineering of China

摘　　要：针对具有不确定特性的大滞后系统,基于传统的Smith预估器结构,提出了一种改进方案。利用相关性分析技术实时调整预估器的滞后时间估计,并采用自适应模糊补偿器在线修正预估器的增益估计,在对象参数不稳定或受扰动的情况下可使预估器的参数迅速跟踪并收敛于对象参数,从而确保系统状态的稳定。该方案增强了系统的鲁棒性,克服了传统Smith预估器需要精确模型的缺点。仿真结果表明,该方案具有较强的参数适应性,对时变大滞后对象具有良好的控制效果。An improved Smith-predictor based on the structure of traditional Smith-predictor for the systems with uncertainty and long dead time is presented.The estimate of time delay is tuned in real-time based on correlation analysis technique,while an adaptive fuzzy compensator is employed to correct the gain estimation on-line.By this means,the parameters of predictor can rapidly track and convergent to the parameters of plant which are unstable or disturbed,and consequently stability of system is guaranteed.This scheme enhances robustness of the controlled system and alleviates the disadvantage of traditional Smithpredictor in which an exactly precise model is needed.Simulation results show that the proposed method has good parametric adaptability and can provide higher characteristics to the controlled objects with time-varying parameters and long dead time.

关 键 词：相关性分析 模糊 SMITH预估器 大滞后 平均幅度差函数 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]