基于支持向量机的煤矸识别研究  

Research on Coal and Gangue Identification Based on Support Vector Machine

作　　者：李黎[1] 熊英[1] 荆瑞俊[2] Li Li;Xiong Ying;Jing Ruijun(Hubei University of Technology, Wuhan Hubei 430068, China;Shanxi Agricultural University, Jinzhong Shanxi 030801, China)

机构地区：[1]湖北工业大学,湖北武汉430068 [2]山西农业大学,山西晋中030801

出　　处：《山西电子技术》2021年第6期85-87,共3页Shanxi Electronic Technology

摘　　要：矸石是影响采煤质量的重要因素,传统的矸石识别主要靠人工,综采面的环境条件极差严重影响了工人的健康;矸石对后期的洗煤工艺也提出了挑战。鉴于此,本文研究基于声音的煤矸识别系统,以提高采煤的质量。本系统首先采集到了采煤过程中存在放顶煤时煤落下、矸石落下的声音及没有放煤的声音,对比学习声音特征提取算法MFCC和logfbank优劣,系统采用logfbank算法实现特征提取;采用支持向量机分类器对得到数据集的特征参数进行训练,构建分类识别模型;用训练得到模型对测试数据进行识别,识别率为88.33,达到工业现场使用的要求。Gangue is an important factor affecting the quality of coal mining.The traditional identification of gangue mainly relies on manual work.The extremely poor environmental conditions of fully mechanized mining face seriously affect the health of workers.In view of this,this paper studies the coal and gangue identification system with sound in order to improve the quality of coal mining.The system first collects the sound of coal falling,gangue falling and no coal pouring,and then makes comparing with MFCC and logfbank algorithm to learn the sound feature extraction.logfbank algorithm is better in the system.Support vector machine classifier is used to train the characteristic parameters of the obtained data set,and the classification and recognition model is constructed.The recognition model is used to identify the test data,and the result is 88.33,which can be used in industrial field.

关 键 词：支持向量机 煤矸石 MFCC 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] TP391.42[自动化与计算机技术—控制科学与工程]