Page 212 - 《振动工程学报》2026年第5期
P. 212
第 39 卷第 5 期 振 动 工 程 学 报 Vol. 39 No. 5
2026 年 5 月 Journal of Vibration Engineering May 2026
振 动 主 动 控 制 双 梯 度 均 衡 算 法 研 究
赵晏萱 , 李新辉 , 杨铁军 , 朱明刚 , 吴 磊 , 徐 阳 2
1
1
1
1
2
(1. 哈尔滨工程大学动力与能源工程学院,黑龙江 哈尔滨 150001; 2. 哈尔滨工程大学青岛创新发展基地,山东 青岛 266000)
摘要:振动主动控制技术基于波叠加原理,利用次级振源产生振动来抵消初级振源的振动,实现“以振制振”的目的。该技术
在工程应用中,当控制输出信号超出系统硬件能力时,会引发输出饱和问题,使得硬件工作在非线性区域,从而导致基于线性
假设的算法控制效果减弱,严重时甚至引起系统失稳和硬件损坏。为解决该问题,基于滤波 x 最小均方(filter-x least mean square,
FxLMS)算法的约束输出主动控制算法被提出,其中以泄露算法和双梯度算法最具代表性。泄露算法因缺乏泄漏因子与输出
功率的对应关系,导致算法参数调节困难;双梯度算法因两个步长因子最优比例关系不明确,导致算法性能受限。本文针对
双梯度算法两个步长因子比例关系问题开展理论分析,提出双梯度均衡控制原则,推导出双梯度算法中两个步长因子的最优
比例关系,提出了双梯度均衡 FxLMS(equalized two-gradient direction FxLMS,ETGD FxLMS)算法,并开展了算法对比仿真分
析,结果表明本文所提算法较双梯度算法性能得到了明显提升。
关键词: 振动主动控制;FxLMS 算法;双梯度 FxLMS 算法;双梯度均衡 FxLMS 算法;约束输出算法
中图分类号:TP13 文献标志码:A DOI:10.16385/j.cnki.issn.1004-4523.202404044
Equalized two-gradient direction algorithm for active vibration control
2
1
1
1
1
ZHAO Yanxuan ,LI Xinhui ,YANG Tiejun ,ZHU Minggang ,WU Lei ,XU Yang 2
(1.School of Power and Energy Engineering,Harbin Engineering University,Harbin 150001,China;
2.Qingdao Innovation and Development Base,Harbin Engineering University,Qingdao 266000,China)
Abstract:Vibration active control technology is based on the principle of wave superposition,which utilizes secondary vibration sources to
generate vibrations to counteract the vibrations of primary vibration sources,achieving the goal of “vibration suppression by vibration”. In
engineering applications, when the control output signal exceeds the hardware capabilities of the system, it can cause output saturation
problems,causing the hardware to operate in a nonlinear region,thereby weakening the control effectiveness of control algorithms based on
linear assumptions,even causing system instability and hardware damage in severe cases. To address this issue,a constrained output active
control algorithm based on the FxLMS algorithm has been proposed, with leaky algorithm and two-gradient algorithm being the most
representative. The leaky algorithm suffers from difficulties in adjusting algorithm parameters due to the lack of correspondence between the
leakage factor and output energy. The two-gradient algorithm suffers from limitations in algorithm performance due to the unclear optimal ratio
relationship between the two step size factors. In this paper, theoretical analysis is conducted to address the issue of the optimal ratio
relationship between the two step size factors in the two-gradient algorithm,a two-gradient equilibrium control principle is proposed,the
optimal ratio relationship between the two step size factors in the two-gradient algorithm is derived,an ETGD-FxLMS algorithm is proposed,
and algorithm comparison simulation analysis is conducted. The results show that the proposed algorithm significantly improves the
performance compared to the two-gradient algorithm.
Keywords:active vibration control;FxLMS algorithm;TGD FxLMS algorithm;ETGD FxLMS algorithm;constrained output algorithm
振动主动控制又被称作振动有源控制,该方法 振动主动控制常用的控制算法之一是滤波 x 最
的主要思想是在被控设备上安装次级振源(如作动 小均方(filter-x least mean square,FxLMS)算法 。该
[4]
器),通过产生一个与初级激励响应等幅反向的振动 算法具有简单易于实现、计算复杂度低、稳定性高
响应,实现振动互相抵消,从而起到“以振制振”的 等优点。但是,该算法并未考虑控制系统硬件输出
目的 。振动主动控制能够适应被控对象工况变化, 能力的限制,当输出信号超出硬件(如数模转换器、
[1]
并具有优异的振动线谱控制能力,近年来得到了越 功率放大器或作动器)能力时,会导致硬件工作在非
来越广泛的工程应用 [2-3] 。 线性区域,这使得基于线性假设的 FxLMS 算法稳定
收稿日期:2024-04-19;修订日期:2024-06-22
基金项目:国家自然科学基金资助项目(52001092)

