摘要
研究了多频波动线谱激励下的主动隔振控制问题,提出一种改进小波包自适应控制算法.传统的自适应算法难以同时控制多线谱,当线谱存在频率波动时,窄带滤波算法难以取得理想的控制效果.因此,将小波包分解算法应用在控制算法中,对信号进行等间距频带划分,采取并行控制策略完成多线谱控制.分析了不同小波基函数对信号分解的影响,确定最优小波基.针对虚有线谱和子频带过多导致的控制过盈问题,改进小波包分解算法,降低了计算量,消除子频带中虚有线谱.为每个子频带设置独立步长,并根据箕舌线函数修正迭代步长.搭建控制系统开展试验研究,结果表明,本文所提改进小波包自适应算法在多频波动线谱激励下能够取得良好的控制效果,线谱振动能量显著降低.
The active vibration isolation control problem under multi⁃frequency fluctuant line spectrum excitation was studied,and an improved wavelet packet adaptive control algorithm was proposed.It was difficult for traditional adaptive algorithms to control multi⁃frequency line spectra at the same time.Narrowband filtering algorithms also could not achieve good effect under fluctuant frequency excitation.Therefore,the wavelet packet decomposition algorithm was introduced to the control algorithm to divide the signal into equally spaced frequency bands.And then,a parallel control strategy was adopted to control multi⁃frequency line spectrum.The influence of different wavelet basis functions on signal decomposition was analyzed,and the optimal wavelet basis was determined.Aimed at the problems caused by the virtual line spectrum and too many sub⁃bands,the wavelet packet decomposition algorithm was improved to reduce the calculation amount and eliminate the virtual line spectrum in the sub⁃band.An independent step size was set for each sub⁃band,and modified according to the tongue⁃line function.The control system was build to carry out experiment.The results showed that the improved wavelet packet adaptive algorithm proposed in this paper can achieve good control effect under multi⁃frequency fluctuant line spectrum excitation,and the vibration energy of line spectrum was significantly reduced.
作者
高伟鹏
贺国
刘树勇
GAO Weipeng;HE Guo;LIU Shuyong(College of Power Engineering,Naval University of Engineering,Wuhan 430033,China)
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2021年第1期168-175,共8页
Journal of Harbin Institute of Technology
基金
国家自然科学基金(51579242)
国家自然科学基金青年科学基金(51509253)。
关键词
振动控制
多频波动线谱
dmey小波基
小波包分解改进算法
变步长
vibration control
multi⁃frequency fluctuant line spectra
dmey wavelet basis
improved wavelet packet decomposition algorithm
variable⁃step
