基于频率相关的压电作动器迟滞特性建模及参数辨识  

Modeling and Parameter Identification of Rate-Dependent Hysteresis of Piezoelectric Actuators

作　　者：钱承 殷小亮 QIAN Cheng;YIN Xiaoliang(Jiaxing university,Mechanical and Electrical Engineering College,Jiaxing Zhejiang 314001,China)

机构地区：[1]嘉兴学院机电工程学院,浙江嘉兴314000

出　　处：《传感技术学报》2020年第5期699-704,共6页Chinese Journal of Sensors and Actuators

基　　金：浙江省自然科学基金青年项目(LQ19A020002);嘉兴市科技局公益项目(2019AD32032)。

摘　　要：为辨识压电作动器在应用过程中的率相关迟滞特性,采用基于最小二乘支持向量机理论对压电作动器进行建模,并引入粒子群算法对模型的参数进行优化。模型以当前及历史输入电压、历史输出位移组成的向量作为模型的输入,而以当前的输出位移作为模型的输出进行训练。最后,通过仿真结果和实验结果对比发现,本文建立的压电率相关迟滞模型平均误差为0.0208μm,最大误差为0.4290μm比Bouc-Wen模型精度高,验证了本文建模方法的可行性。In order to identity rate-dependent hysteresis characteristics of piezoelectric actuator in the process of application,least square support vector mechanism theory is adopted and parameters of LS-SVM are optimized by the particle swarm algorithm.Taking current and history of input voltage,history of output displacement vector as input of the model,and taking current output displacement as the output of the model for training.Finally,by comparison with the results of simulation and experiment results,the rate-dependent hysteresis LS-SVM model is better than Bouc-Wen model,and the average error is 0.0208μm and the maximum error is 0.4290μm of the rate-dependent hysteresis model.Thus,the feasibility of LS-SVM method in this paper is verified.

关 键 词：压电作动器 率相关迟滞特性 最小二乘支持向量机 参数辨识 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]