摘要
在高频响应下,音圈电机存在明显的迟滞现象,并且具有一种特殊的复杂迟滞特性,即具有动态、非光滑和非单调的复杂迟滞特性。为了有效地描述复杂迟滞特性,提出了基于构建类迟滞的高频响音圈电机复杂迟滞混合神经网络建模方法,可实现复杂迟滞模型参数的自适应调整。实验结果表明,迟滞混合建模方法能有效地描述高频响音圈电机复杂迟滞特性以及对应的复杂迟滞的逆,并且模型具有较高的精度。
In high frequency response, the hysteresis phenomenon for voice coil motor is a complex behavior of a dynamic and non - smooth and non - monotonic. In order to concisely describe the complex hysteresis behavior, the new modeling method for the complex hysteresis was proposed, in which the neural networks was integrated with the constructed similar hysteresis model by which input signals are processed. Parameters in the new modeling method can be adjusted adaptively. The experiment results show the effectiveness of proposed modeling method.
出处
《计算机仿真》
CSCD
北大核心
2012年第12期219-223,282,共6页
Computer Simulation
基金
国家自然科学基金项目(60964001)
广西自然科学基金重点项目(桂科自0991019Z)
广西信息与通讯技术实验室基金项目(10902)
关键词
非单调迟滞特性
类迟滞
音圈电机
迟滞混合建模
神经网络
Non - monotone hysteresis behavior
Similar hysteresis
Voice coil motor
Hysteresis hybrid modeling
Neural networks
