超磁致伸缩执行器磁滞模型的参数辨识  

Parameter Identification of Hysteresis Model for Giant Magnetostrictive Actuator Based on Jiles-Atherton Model

作　　者：唐宏波[1] 朱玉川[1] 

机构地区：[1]南京航空航天大学机电学院,江苏南京210016

出　　处：《压电与声光》2015年第5期863-866,872,共5页Piezoelectrics & Acoustooptics

基　　金：国家自然科学基金资助项目(51175243);航空科学基金资助项目(20130652011);江苏省自然科学基金资助项目(BK20131359)

摘　　要：准确辨识磁滞模型参数是保证超磁致伸缩执行器位移控制精度的关键,而单一算法难以实现对超磁致非线性模型参数的精确辨识。该文提出了一种新型混合优化策略,即改进的遗传退火算法,并将其应用于对超磁致伸缩执行器位移磁滞模型参数的辨识。该算法兼顾了遗传算法和模拟退火算法的优点,同时还引入了机器学习原理,将模拟退火算法作为遗传算法中的种群变异算子,并将模拟退火算法中的抽样过程与遗传算法相结合。此算法不仅充分发挥了遗传算法并行搜索能力强的特点,且增强和改进了遗传算法的进化能力,同时提高了系统的收敛性和收敛速度,避免最优解的丢失。通过仿真和试验研究表明,该算法相对于遗传算法有更高的精度,可有效精确辨识超磁致伸缩执行器磁滞模型的参数。Accurately identifying the hysteresis model parameters may improve the control precision of giant magnetostrictive actuator output displacement.The single algorithm is difficult to achieve accurate identification of the ultra-magnetically induced nonlinear model parameters.This paper presents a new hybrid optimization strategy,the modified genetic algorithm and simulated annealing and applied to identification of the giant magnetostrictive actuator displacement hysteresis model parameters.The algorithm taking into account the genetic algorithm and simulated annealing algorithm strengths,and also introduces machine learning theory,simulated annealing algorithm as the population variation in genetic algorithm operator and simulated annealing algorithm and genetic algorithm Metropolis sampling process combines.This algorithm not only give full play to the ability of genetic algorithms parallel search features,but also enhance and improve the ability of genetic algorithms evolution and improve the convergence and convergence speed of the system,to avoid losing the optimal solution.The simulation and experimental results show that the algorithm with respect to the genetic algorithm has a high accuracy,the parameters can effectively identify the model.

关 键 词：Jiles-Atherton磁滞模型 超磁致伸缩执行器 遗传算法 改进的遗传退火算法 参数辨识 

分 类 号：TM153[电气工程—电工理论与新技术] V227.83[航空宇航科学与技术—飞行器设计]