摘要
对于现有减法平均优化(SABO)算法在寻优变分模态分解(VMD)的参数惩罚因子(α)和分解模态数(K)时存在的不足,提出改进减法平均优化(GSABO)算法。为克服SABO算法存在的不足,初始化粒子时用混沌映射代替随机数来增加粒子多样性,在迭代寻优的过程中,用黄金正弦辅助粒子跳出局部最优解。针对当前硬阈值和软阈值函数降噪时,前者在阈值点处是非连续的,而后者存在固定偏差的缺点。提出改进小波阈值,然后通过VMD去噪联合改进小波阈值去噪对西储大学轴承数据进行去噪研究,并与VMD、传统小波阈值和改进小波阈值的去噪性能进行比较。结果表明:GSABO优化VMD去噪联合改进小波阈值去噪拥有更佳的效果,在保留有效信号的同时,能更加提高信噪比和减少均方根误差。
Aiming at the shortcomings of the existing subtractive average optimization(SABO)algorithm in optimizing the parameter penalty factor(α)and the decomposition mode number(K)of variational modal decomposition(VMD),an improved subtractive-average-based optimizer(GSABO)algorithm is proposed.In order to overcome the shortcomings of SABO algorithm,chaotic mapping is used instead of random number to increase the diversity of particles when initializing particles,and golden sine is used to help particles jump out of local optimal solution during iterative optimization.At present,the hard threshold and soft threshold function are discontinuous at the threshold point,while the latter has a fixed deviation.An improved wavelet threshold is proposed,and then the bearing data of Case Western Reserve University are denoised by VMD denoising combined with improved wavelet threshold denoising,and the denoising performance is compared with VMD,traditional wavelet threshold and improved wavelet threshold.The results show that GSABO optimized VMD denoising combined with improved wavelet threshold denoising has better effect,which can improve the signalto-noise ratio and reduce the root mean square error while retaining effective signals.
作者
金志浩
魏新志
JIN Zhihao;WEI Xinzhi(School of Mechanical and Power Engineering,Shenyang University of Chemical Technology,Shenyang 110142,Liaoning,China)
出处
《中国工程机械学报》
2025年第5期891-895,912,共6页
Chinese Journal of Construction Machinery
基金
国家自然科学基金联合基金资助项目(U1708254)
国家自然科学基金青年科学基金资助项目(11702178)。
关键词
改进减法平均优化算法
变分模态分解
改进小波阈值函数
贝叶斯阈值
轴承去噪
improved subtraction average optimization algorithm
variational modal decomposition
improved wavelet threshold function
Bayesian threshold
bearing denoising
