摘要
针对化工生产中利用放射源穿透物料进行密度检测时,由于光电倍增管输出的光电转换信号中夹杂高斯白噪声而导致有效信号提取困难的问题,提出了一种自适应完备噪声经验模态分解(complete ensemble empirical mode decomposition with adaptive noise,CEEMDAN)与改进小波阈值相结合的去噪方法。首先构建光电转换信号函数模型,然后对信号进行模态分解,对分解后的模态函数通过相关系数法进行筛选,之后采用改进的小波阈值函数结合模拟退火优化阈值参数,使其更为适合信号函数模型,从而去除本征模态函数中的噪声成分,最后重构得到去噪后信号。通过试验对比表明,该方法能有效降低噪声干扰,在信噪比、均方根误差、波形相似参数均优于对比算法,提高了检测的精度。
In response to the challenge of extracting effective signals from the photoelectric conversion signals output by a photomultiplier tube,which is contaminated by Gaussian white noise during density measurement using radioactive sources in chemical production,this study proposes a denoising method that integrates adaptive complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)and an improved wavelet thresholding approach.First,a functional model of the photoelectric conversion signal was constructed.The signal was then subjected to modal decomposition,and the resulting intrinsic mode functions were screened using the correlation coefficient method.Subsequently,an improved wavelet threshold function was employed,with threshold parameters optimized via simulated annealing to better fit the signal model,thereby removing noise components from the intrinsic mode functions.Finally,the denoised signal was reconstructed.Experimental comparisons demonstrate that the proposed method effectively reduces noise interference,achieving superior signal-to-noise ratio,root mean square error,and waveform similarity parameters compared to baseline algorithms,thereby enhancing measurement accuracy.
作者
王小鹏
许金玉
李家隆
WANG Xiaopeng;XU Jinyu;LI Jialong(School of Electronic and Information Engineering,Lanzhou Jiaotong University,Lanzhou 730070,China)
出处
《振动与冲击》
北大核心
2025年第24期278-288,共11页
Journal of Vibration and Shock
基金
甘肃省高等学校产业支撑计划项目(2023CYZC-40)
甘肃省优秀研究生“创新之星”项目(2025CXZX-685)。
