基于分数信号处理的变步长自适应振动主动控制方法研究

New variable step strategy for online secondary path modeling with fractional signal processing approach in active vibration control systems

针对自适应振动主动控制系统中次级通道的辨识精度严重影响振动控制效果的问题,分析了常规的次级通道在线辨识算法存在的问题,提出一种基于分数信号处理的双步长两阶段变步长策略的次级通道在线辨识方法。该方法使用基于分数信号处理的自适应算法代替传统的最小均方(least mean square, LMS)算法进行次级通道的在线辨识,同时给出了一种双步长的两阶段变步长策略,在次级通道辨识环节收敛前后应用不同的变步长策略以提高辨识精度和降低辨识环节的波动。仿真结果表明,与现有方法比较,该方法的次级通道辨识收敛速度更快,系统收敛后的波动更小,次级通道的辨识精度和系统的稳定性都有了明显的提升。经验证,该方法有效解决了常规的次级通道在线辨识算法收敛速度慢、辨识精度低和辨识环节波动大等问题,具有更好的振动控制效果。

Aiming at the problem that the precision of identification of the secondary path in the adaptive vibration active control system seriously affects the performance of vibration control, this paper analyzed the problem existing in the conventional online identification algorithm of the secondary path and proposed an online identification method based fractional signal processing. The adaptive strategy of fractional signal processing replaced the traditional least mean square algorithm and presented a two step strategy consisting of two stage variable step size. In order to improve the precision and reduce the fluctuation of the identification, different variable step size strategies are used before and after the convergence of the secondary path identification procedure. The simulation results show that, compared with the existing methods, the proposed method can significantly improve the identification convergence speed of the secondary path, effectively avoid the fluctuation of the identification after the convergence of the identification, and improve the identification precision of the secondary path and the stability of the system. It is verified that this method can effectively solve the problems of slow convergence, low identification accuracy and large fluctuation of identification part in the conventional online identification algorithm of the secondary path and a better vibration control performance can be achieved.