Giant magnetostrictive material (GMM), as a new type of functional material, was widely used in energy harvesting, micro displacement driving, precision positioning control and other fields due to its advantages of large magneto mechanical coupling coefficient, fast response speed, and good frequency response characteristics. However, the complex hysteresis nonlinearity of the material affected the positioning accuracy of its actuator. In order to identify the hysteresis nonlinearity in GMM materials, this paper proposed a new Hammerstein model modeling method. The advantage of this method was that the model could better approximate hysteresis nonlinearity, provide higher accuracy, and reduce the workload of parameter identification in the series link. Firstly, an extreme learning machine model was constructed based on hyperbolic functions to represent the static nonlinear part of the new Hammerstein model in the extended space of hysteresis operators. Secondly, the extracted weights and bias parameters of the fully connected layers of the extreme learning machine model was used to construct the state space equation of the dynamic linear part in the new model, which reduced the workload of identifying model parameters in the traditional model with serial links. Finally, a new Hammerstein model was established to describe the hysteresis characteristics of giant magnetostrictive materials. The modeling relative error percentage of the new Hammerstein model is 0.86% to 3.69%, and the average absolute error percentage is 2.63%, which is about 0.8 lower than the root mean square error of the traditional Hammerstein model, and the average absolute percentage error increases 4%. The simulation results demonstrate the effectiveness of the new Hammerstein model in modeling the complex hysteresis characteristics of giant magnetostrictive materials.
1) 静态非线性环节的建模方法研究, 研究的重点在于提高精度以及泛用性。具体可以分为三类: ① 使用基函数(如Sigmoid函数等)的线性组合, 缺点是参数多、 阶数高; ② 采用非参数的形式, 即多项式形式[14], 但是不适合于分段非线性问题; ③ 采用模糊系统、 神经网络等非线性模型[15]建立非线性环节, 存在非线性参数化复杂的问题。
2) Hammerstein模型中串联环节带来的新问题, 以及由于中间变量无法测量造成的静态非线性环节和动态线性环节组合的辨识难度提高的问题。研究重点在于模型参数求解的问题。主要有两种解决方法: ① 采用迭代法、 随机法、 频域法及盲辨识法等方法去辨识串联参数; ② 使用组合信号、 改进辨识算法[16-17]等方法进行静态非线性环节和动态线性环节的参数的辨识。
HEXinghui, WANGZiwei, CHENBingqian. Analysis of the application and development of piezoelectric ceramic materials[J].Create Living, 2018, 10(5): 1.(in Chinese)
CHENGMei, QIUYe, LIUXuhui, et al. Research on giant magnetostrictive drive and its micro displacement characteristics[J]. Machine Tool & Hydraulics, 2020, 48(8): 78-80.(in Chinese)
[5]
KIMB, WASHINGTONG N, YOONH S. Hysteresis-reduced dynamic displacement control of piezoceramic stack actuators using model predictive sliding mode control[J]. Smart Materials and Structures, 2012, 21(5): 055018.
LIUJie, ANKun, WANGYafeng, et al. Modeling of preisach hysteresis model based on neural networks[J]. Modular Machine Tool & Automatic Manufacturing Technique, 2022(3): 5-8.(in Chinese)
PANMingjian, ANKun, LIJianhong, et al. Nonlinear control of giant magnetostriction based on CMAC neural network[J]. Electronic Measurement Technology, 2023, 46(9): 182-188.(in Chinese)
[11]
MielkeA. Generalized Prandtl⁃Ishlinskii operators arising from homogenization and dimension reduction[J]. Physica B: Condensed Matter, 2012, 407(9): 1330-1335.
[12]
ZHUW, RUIX T. Hysteresis modeling and displacement control of piezoelectric actuators with the frequency-dependent behavior using a generalized Bouc-Wen model[J].Precision Engineering, 2016, 43: 299-307.
[13]
JIJ A, ZHAOZ G, ZHANGS, et al. Hysteresis Characteristics prediction method of nanocrystalline materials under high-frequency excitation based on Maxwell’s equation and R-L fractional derivative[J/OL]. Proceedings of the CSEE, (2023-08-07)[2024-01-08]
FANJiahua, MALei, ZHOUPan, et al. Modeling and control of piezoelectric actuators based on radial basis function neural network[J]. Control Theory & Applications, 2016, 33(7): 856-862.(in Chinese)
[18]
NAJ, CHENQ, RENX. Adaptive identification and control of uncertain systems with non-smooth dynamics[M]. Pittsburgh:Academic Press, 2018.
SONGWei, HANJiahu, LIFeng, et al. Identification of Hammerstein nonlinear model under colored noise interference[J]. Journal of Shaanxi University of Science and Technology, 2023, 41(5): 189-194.(in Chinese)
[23]
HEH R, NAJ, WUJ D, et al. Fixed-time adaptive parameter estimation for hammerstein systems subject to dead-zone[J]. IEEE Transactions on Industrial Electronics, 2023, 71(4): 3862-3872.
[24]
PAPADIMITRIOUC, VARELMANNT, SCHRÖDERC, et al. Globally optimal scheduling of an electrochemical process via data-driven dynamic modeling and wavelet-based adaptive grid refinement[J]. Optimization and Engineering, 2023: 1-39.
[25]
LÜL, SUNW, PANJ. Two‐stage and three‐stage recursive gradient identification of Hammerstein nonlinear systems based on the key term separation[J]. International Journal of Robust and Nonlinear Control, 2024, 34(2): 829-848.
[26]
HUANGG B, ZHUQ Y, SIEWC K. Extreme learning machine: Theory and applications[J]. Neurocomputing, 2006, 70(1/2/3): 489-501.
WANGYan, ZHUWei, WANGJunliang, et al. End analysis of flexible FBG shape reconstruction based on ELM algorithm[J]. Chinese Journal of Scientific Instrument, 2023, 44(5): 81-89.(in Chinese)