摘要
提出了基于改进最优阈值估计和改进阈值函数两方面优化的降噪方法。首先,针对传统阈值法估计偏大和单序列样本熵阈值法估计易陷入局部最优的问题,建立了融合噪声序列样本熵和降噪序列样本熵的综合熵模型,以综合熵曲率拐点下的阈值为优化目标,提出一种二分变步长非线性搜索方法实现最优阈值的快速估计。进一步,针对传统小波阈值函数存在重构振荡和重构偏差的问题,提出了一种融合软、硬阈值特征的改进阈值函数,通过使用平滑过渡数模型对该函数中的品质因子参数进行优化,使函数曲线在临界阈值邻域内连续平滑过渡。在高压设备局部放电超声波脉冲信号降噪应用场景中的仿真及测试试验结果表明,所提出综合熵阈值估计方法能够快速、准确地逼近最优阈值。此外,改进阈值函数兼顾了软、硬阈值函数的优点,能够抑制重构振荡和重构偏差,在保留原始信号有效信息的前提下实现对噪声的有效抑制,表现出了较好的工程应用价值。
A denoising method was proposed based on two optimization measures, namely, the improved optimal threshold estimation and improved threshold function. In view of the problems that the estimated value of the traditional threshold method is inclined to be relatively large, and the estimation by the single sequence sample entropy threshold method is easy to fall into a local optimum, a comprehensive entropy model was established by fusing the sample entropies of both the noise sequence and noise reduction sequence. Choosing the threshold at the comprehensive entropy curvature inflection point as the optimization target, a bipartite variable step-size nonlinear search method was proposed to realize the rapid estimation of the optimal threshold. Then, aiming at the reconstruction oscillation and reconstruction deviation of the traditional wavelet threshold function, an improved threshold function combining soft and hard threshold features was proposed. To achieve the continuous smooth transition of the function curve in the critical threshold neighborhood, the quality factor parameters in this function were optimized by using a smooth transition number model. The simulation and experiment results in the application of the noise reduction method to the partial discharge ultrasonic pulse signal of a high-voltage equipment show that the proposed comprehensive entropy threshold estimation method can approximate the optimal threshold quickly and correctly. Besides, the improved threshold function also intrgrates the advantages of soft and hard threshold functions, which can suppress reconstruction oscillation and deviation, and further suppress the noise effectively with the premise of retaining original signal’s effective information, showing a good engineering application value.
作者
刘志坚
赵浩益
刘航
罗灵琳
宋琪
李鹏程
LIU Zhijian;ZHAO Haoyi;LIU Hang;LUO Linglin;SONG Qi;LI Pengcheng(Faculty of Electric Power Engineering,Kunming University of Science and Technology,Kunming 650500,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2022年第22期1-13,共13页
Journal of Vibration and Shock
基金
昆明理工大学自然科学研究基金(KKZ3202004042)。
关键词
信号降噪
阈值估计
阈值函数
综合熵
局部放电
signal noise reduction
threshold estimation
threshold function
comprehensive entropy
partial discharge
