[1]刘 艺,杜雨洺,官 毓,等.基于频域的声信号计权改进算法[J].成都信息工程大学学报,2021,36(01):68-72.[doi:10.16836/j.cnki.jcuit.2021.01.011]
LIU Yi,DU Yuming,GUAN Yu,et al.Improved Algorithm of Acoustic Signal Weighting based on Frequency Domain[J].Journal of Chengdu University of Information Technology,2021,36(01):68-72.[doi:10.16836/j.cnki.jcuit.2021.01.011]
点击复制
LIU Yi,DU Yuming,GUAN Yu,et al.Improved Algorithm of Acoustic Signal Weighting based on Frequency Domain[J].Journal of Chengdu University of Information Technology,2021,36(01):68-72.[doi:10.16836/j.cnki.jcuit.2021.01.011]
基于频域的声信号计权改进算法
成都信息工程大学学报[ISSN:1006-6977/CN:61-1281/TN]
卷:
36
期数:
2021年01期
页码:
68-72
栏目:
电子信息科学与技术
出版日期:
2021-02-28
- Title:
- Improved Algorithm of Acoustic Signal Weighting based on Frequency Domain
- 文章编号:
- 2096-1618(2021)01-0068-05
- 关键词:
- 声级计; 时间计权; 频域; 1/3倍频程; Parseval定理
- 分类号:
- TB52
- 文献标志码:
- A
- 摘要:
- 现有高精度数字声级计产品对测量精度和运算速度的要求越来越高,以及对实时同步声压级参数计算的数量需求越来越多。在对声信号进行计算处理时,模拟式声级计在时域下先频率计权,再进行时间计权的传统算法,不仅难以满足现代高精度数字声级计产品高精度和高速度的需要,而且计算参数单一,难以实现计权倍频程声压级等参数的同步实时计算。故提出一种基于Parseval定理和FFT谱分解,先时间计权,再频率计权的改进算法。将声信号做FFT谱分解后,在频域下对频谱信号的功率值进行时间计权和频率计权,最后通过能量求和等方法,实现频率、时间计权总声压级的计算和计权声压级的倍频程分析,以实现计算结果的多参数化。通过与传统算法结果的对比分析,验证算法的有效性与简易性,且满足声级计产品相关标准的基本要求。
- Abstract:
- Existing high-precision digital sound level meter products have higher and higher requirements for measurement accuracy and calculation speed,as well as the number of real-time simultaneous sound pressure level parameter calculations.When calculating and processing sound signals, the traditional algorithm of analog sound level meters in the time domain for frequency weighting and time weighting is not only difficult to meet the high precision and high speed of modern high-precision digital sound level meters. Need, and the calculation parameters are single,it is difficult to realize the simultaneous real-time calculation of the weighted octave sound pressure level and other parameters.Therefore,an improved algorithm based on Parseval’s theorem and FFT spectrum decomposition is proposed,which is time-weighted first,and then frequency-weighted. After the acoustic signal is decomposed by the FFT spectrum, the power value of the spectrum signal is time-weighted and frequency-weighted in the frequency domain. Finally, the calculation and calculation of the frequency and time-weighted total sound pressure level are realized through methods such as energy summation.The octave band analysis of the weighted sound pressure level to realize the multi-parameterization of the calculation results.The comparison and analysis with the results of traditional algorithms verify the effectiveness and simplicity of the algorithm and meet the basic requirements of the relevant standards of sound level meters.
参考文献/References:
[1] 周星,张晨,裴真,等.全数字式声级计系统的设计与实现[J].信号处理,2005,21(S1):609-612.
[2] 陈剑林,白滢,牛峰,等.声级计的频率计权特性[J].计量技术,2008(6):47-50.
[3] 陈剑林,帅正萍,白滢,等.声级计的时间计权特性[J].计量技术,2007(7):44-46.
[4] 王东生.分贝的由来和1/3倍频程算法[J].橡塑资源利用,2005(5):45-46.
[5] Oppenheim A V.信号与系统[M].北京:电子工业出版社,2002
[6] 何岭松.频率计权网络的公式化表达[J].仪器仪表学报,1996,17(5):540-542.
[7] 李靖.数字声级计中Ⅰ计权算法设计与本底噪声估计技术研究[D].长沙:湖南大学,2016.
[8] 成峰,周宝焜,程利青,等.基于MATLAB的13倍频程FIR数字滤波器设计[J].福州大学学报(自然科学版),2003(2):160-165.
[9] 王孚懋,王建春.噪声测量中1/3倍频程与倍频程频谱的关系与计算[J].噪声与振动控制,1996(3):39-41.
[10] 张登攀,高志强.噪声1/3倍频程计权声级算法[J].河南理工大学学报(自然科学版),2013,32(6):709-712.
[11] 吴胜举,张明铎.声学测量原理与方法[M].北京:科学出版社,2014.
备注/Memo
收稿日期:2020-09-25
更新日期/Last Update:
2021-02-28