基于LMF算法的主动控制次级通道传递函数辨识方法研究

Research on Identification Method of Secondary Channel Transfer Function in Active Control Based on LMF Algorithm

摘要: 结构振动的主动控制算法中，次级作动器产生的控制信号要经过传递路径才能到达误差传感器，存在次级通道的传递函数，影响自适应主动控制系统的性能，使系统的稳定性下降甚至失稳。FxLMS算法在有准确误差通道传递函数参与计算时可以有效解决次级通道带来的问题。本文摒弃传统的LMS在线辨识方法，提出了一种基于LMF算法的次级通道在线辨识方法，使用LabVIEW编程语言设计次级通道在线辨识的仿真计算程序，进行仿真实验，对比了LMS算法和LMF算法在辨识滤波器方面的性能。结果表明，当主动控制系统的信噪比大于0时，LMF算法相较于常用的LMS算法，具有收敛速度快、结果稳定性强、辨识传递函数准确的优点，可以作为一种更加适用的算法应用于实际的主动控制系统中。

Abstract: Active control algorithms for structural vibration control encounter problems with secondary channels that involve the transmission of control signals through a transmission path to error sensors, resulting in the existence of secondary channel transfer functions that affect the performance of adaptive active control systems, causing instability and a decrease in stability. The FxLMS algorithm effectively solves the problems caused by secondary channels when there is accurate error channel transfer function involved in calculations. This paper proposes an online secondary channel identification method based on the LMF algorithm, which abandons the traditional LMS online identification method. Using the LabVIEW programming language, a simulation program for online identification of secondary channels is designed and simulated experiments are conducted to com-pare the performance of LMS and LMF algorithms in filter identification. The results show that when the signal-to-noise ratio of the active control system is greater than zero, the LMF algorithm has the advantages of fast convergence speed, strong result stability, and accurate identification of the transfer function compared to the commonly used LMS algorithm, and can be used as a more applicable algorithm for practical active control systems.