基于MFCC的路基土密实度识别技术研究

Abstract
Figure/Table
References (0)
Related Citation (15)

Download: PDF (1896 KB) HTML (1 KB)
Export: BibTeX | EndNote (RIS)

Abstract Research purposes: Most of the characterization indexes used in the outdoor subgrade soil continuouscompaction technology are converted from the frequency and amplitude of the vibration acceleration signal collected on the steel wheel of the vibratory roller, and the characterization effect is poor under the influence of jumping vibration andmechanical natural vibration, and the signal has not been deeply excavated. In order to find reliable indicators tocharacterize soil compaction degree, this paper introduces sound source material recognition technology while conductingindoor test and analysis, and extracts timbre features based on human ear hearing and psychology, combined withphysical properties of materials, so as to improve the recognition rate of soil compaction degree. By stimulating soil samples with different compacted degrees indoors, acceleration signals and sound signals were collected. MIRtoolbox timbre toolbox was introduced to extract sound timbre features using the impact sound feature extraction method based on the principle of hearing perception, and this method was applied to the feature extraction of acceleration signals.
Research conclusions: (1) When the hammering force is 55 N± 5 N, the correlation coefficients between theeigenvalue mfcc6max extracted from the acceleration and sound signal and the soil sample compaction degree are 0.81 and 0.69 respectively, so the eigenvalue mfcc6max can be used as the index to characterize the soil compaction degree.(2) In the reverse identification of soil compaction, the hammering force is controlled at 55 N± 5 N, and the mfcc6max value calculated in the acceleration signal is substituted into the correlation equation y=0.014x-0.92, or the mfcc6max value calculated in the sound signal is substituted into the correlation equation y=0.018x-1.09. The two obtained y values can predict the soil compaction degree in reverse.(3) The characteristic values and methods used to characterizesoil compaction degree in this test can be further applied to outdoor continuous compaction technology, and are fundamentally different from traditional continuous compaction technical indicators, which provide guidance for improving soil recognition rate.

Key words： MFCC frequency band subgrade compactness hammering force rank correlation analysis

Received: 26 January 2022

PACS:

U416.04

	Service

	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	GUO Huajie
	JIANG Huihuang
	LIU Xiangming

Cite this article:

GUO Huajie,JIANG Huihuang,LIU Xiangming. Research on the Identification Technology of Subgrade Soil Density Based on MFCC[J]. Journal of Railway Engineering Society, 2023, 40(11): 8-16.

URL:

https://tdgcxb.crec.cn/EN/ OR https://tdgcxb.crec.cn/EN/Y2023/V40/I11/8

[1]	J. C. Brown.Musical Instrument Identification Using Pattern Recognition with Cepstral Coefficients as Features[J].Journal of the Acoustical Society of America, 1999,105(3):1933-1941.
[2]	董滨,赵庆卫,颜永红.基于共振峰模式的汉语普通话中韵母发音水平客观测试方法的研究[J].声学学报,2007(2):122-128.Dong Bin,Zhao Qingwei,Yan Yonghong.Objective Evaluation of Vowels of Standard Chinese Pronunciation Based on Formant Pattern [J].Acta Acoustica,2007(2):122-128.
[3]	王娜.基于人耳主管反应的听觉特征量及其在目标识别中的应用[D].西安:西北工业大学,2006.Wang Na. Auditory Features Extraction Based on Human Subjective Response and Applications into the Target Recognition[D]. Xi'an:Northwestern Polytechnical University,2006.
[4]	王娜,陈克安,黄凰.水下噪声听觉属性的主观评价与分析[J].物理学报,2009(10):7330-7338.Wang Na,Chen Kean,Huang Huang.Subjective Evaluation and Analysis of Auditory Attributes for Underwater Noise[J].Acta Physica Sinica,2009(10):7330-7338.
[5]	梁雍. 冲击声的声源特性表达及自动识别研究[D]. 西安:西北工业大学,2017.Liang Yong. Research on Sound Source Characteristic Expression and Automatic Recognition of Impact Sound [D]. Xi'an: Northwestern Polytechnical University,2017.
[6]	赵凯,董明明,刘锋,等.基于声信号的履带机器人地面分类试验研究[J].北京理工大学学报,2018(9):912-916.Zhao Kai,Dong Mingming,Liu Feng,etc.Experimental Study on Terrain Classification Based on Acoustic Signa for Tracked Robot[J].Journal of Beijing Institute of Technology,2018(9):912-916.
[7]	叶阳升,蔡德钩,朱宏伟,等.基于振动能量的新型高速铁路路基压实连续检测控制指标研究[J].铁道学报,2020(7):127-132.Ye Yangsheng, Cai Degou, Zhu Hongwei,etc. Research on New Continuous Compaction Control Index of High-speed Railway Subgrade Based on Vibration Energy [J]. Journal of the China Railway Society, 2019(7):127-132.
[8]	张庆龙. 土石方压实监控系统及其应用研究[D].北京:清华大学,2018.Zhang Qinglong. Study on the Compaction Monitoring System of Earthwork and Its Application[D]. Beijing:Tsinghua University, 2018.