对压电陶瓷执行器迟滞特性的智能建模  

Intelligent Modeling for Hysteresis Nonlinearity in Piezoceramic Actuator

作　　者：赵新龙[1] 谭永红[2] 

机构地区：[1]上海交通大学电子信息学院自动化系,a0303222上海200030 [2]桂林电子工业学院智能系统与工业控制研究室,桂林541004

出　　处：《系统仿真学报》2006年第1期23-25,共3页Journal of System Simulation

基　　金：国家自然科学基金(50265001)

摘　　要：由于压电陶瓷执行器固有的迟滞非线性会影响系统的控制精度,甚至能导致系统不稳定,基于神经网络建立一个迟滞非线性的智能模型。提出一个迟滞因子将多映射的迟滞非线性转换成能够被神经网络逼近的一一映射。该模型结构简单,简化了辨识过程。最后用该方法对实际的压电陶瓷执行器进行建模,结果表明,该模型能够准确的逼近迟滞非线性。A major limitation of piezoceramic actuators is their lack of accuracy due to the hysteresis nonlinearity which can result in the degradation of system performance and even lead to instability. A hysteresis factor wasproposed to transform the multi-valued mapping of hysteresis into a one-to-one mapping which enabled neural networks to approximate the behavior of hysteresis. The proposed model based on neural networks has a simple architecture and simplifies identification procedure. Results illustrate that the proposed method can approximate the hysteresis nonlinearity accurately.

关 键 词：迟滞 PREISACH模型 迟滞因子 智能建模 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]