作者：凌菁, 万江华, 胡甜甜
作者单位：宁夏大学物理与电子电气工程学院，宁夏 银川 750021
关键词：掺杂煤炭;交流阻抗;支持向量机;交叉验证;网格搜索
摘要：
针对传统煤炭掺杂率实验室分析法工序复杂、检测效率低的不足，提出将交流阻抗技术与软测量技术相结合的煤炭掺杂率软测量方法。以宁夏无烟煤为试样验证煤炭交流阻抗特性，提取试样交流阻抗测量值和水分含量为辅助变量，构建基于网格搜索和交叉验证支持向量机（gs-cv svm）的煤炭掺杂率软测量模型，结合交叉验证和网格搜索两种策略优化模型参数。建立仿真实验平台以验证方法有效性，配制掺杂率0~45%的煤炭试样，测量其交流阻抗值和水分含量完成掺杂率识别实验。测试结果表明对于掺杂率在0~45%范围内的煤炭试样，该文方法的准确识别率为85%及以上。
application of support vector machine based on grid search and cross validation in soft-sensing of adulated coal
ling jing, wan jianghua, hu tiantian
school of physics and electronic-electrical engineering, ningxia university, yinchuan 750021, china
abstract: most of the traditional coal adulteration detection methods are laboratory analysis methods with complicated procedures and low detection efficiency. to address this problem, a soft-sensing method for coal adulteration rate is proposed by combining ac impedance technology with soft-sensing technology. the ac impedance characteristics of coal are verified by using ningxia anthracite coal as the sample, a soft measurement model of coal adulteration rate based on grid search and cross-validation support vector machine (gs-cv svm) is established. the ac impedance value and moisture content of coal samples are used as auxiliary variables. the model parameters are optimized by combining two strategies of cross-validation and grid search. the simulation platform is established to verify the validity of the method. coal samples with adulteration rates of 0-45% are prepared, and their ac impedance values and moisture contents are measured to complete the adulteration rate identification experiments. the test results showed that the identification rate is more than 85% for adulated coal samples with adulteration rate of 0-45%.
keywords: adulated coal;ac impedance;svm;cross validation;grid search
2022, 48(8):34-39 收稿日期: 2022-05-02;收到修改稿日期: 2022-06-17
基金项目: 宁夏自然科学基金资助项目（2020aac03008）
作者简介: 凌菁（1982-），女，宁夏银川市人，副教授，硕士生导师，研究方向为智能检测与信息融合
参考文献
[1] 刘福国, 崔福兴. 电厂锅炉入炉煤采制样不确定度分析与模拟[j]. 计量学报, 2019(4): 603-609
[2] 徐水秀, 喻子彧, 覃淮青, 等. 基于激光诱导击穿光谱的煤质快速分析研究及应用[j]. 量子电子学报, 2021, 38(6): 727-750
[3] mao j j, song z l. research on the on-line microwave measurement for moisture in coal and the tests of the influence factors[j]. process automation instrumentation, 2014, 35(7): 22-24
[4] wang w, wang y, shi g. experimental investigation on the infrared refraction and extinction properties of rock dust in tunneling face of coal mine[j]. applied optics, 2015, 54(35): 10532-10540
[5] 史美伦. 交流阻抗谱原理及其应用[m]. 北京: 国防工业出版社, 2001.
[6] 程园新, 韩楠楠, 丁恩杰, 等. 基于有效介质理论的煤岩界面识别方法研究[j]. 煤炭科学技术, 2017, 45(10): 123-128
[7] 林海军, 李松辉, 汪鲁才, 等. 基于交流阻抗的掺杂煤炭快速识别方法[j]. 电子测量与仪器学报, 2015, 29(4): 577-583
[8] 刘陵玉, 杨传法, 张献生, 等. 太赫兹波段煤的湿度与介电特性关系[j]. 煤炭学报, 2016, 41(2): 497-501
[9] li z, bao s, peng x, et al. fault detection and diagnosis in multivariate systems using multiple correlation regression[j]. control engineering practice, 2021, 116(11): 104916
[10] 郝亮, 郭立新. 汽车行驶状态无迹卡尔曼软测量[j]. 中国工程机械学报, 2019, 17(6): 471-475
[11] 王印松, 陈瑞杰. 基于elastic net特征变量选择的scr入口nox软测量模型[j]. 中国测试, 2021, 47(12): 79-86
[12] 谭浩艺, 陈绍炳, 周自强. 基于pca—bp神经网络的锅炉煤质的软测量[j]. 能源技术, 2009(1): 9-11
[13] 郭李娜, 樊贵盛. 基于网格搜索和交叉验证支持向量机的地表土壤容重预测[j]. 土壤通报, 2018, 49(3): 512-518
[14] 蔡波, 黄晋英, 杜金波, 等. 基于meemd多特征融合与ls-svm的行星齿轮箱故障诊断[j]. 中国测试, 2021, 47(9): 126-132
[15] 纪昌明, 周婷, 向腾飞, 等. 基于网格搜索和交叉验证的支持向量机在梯级水电系统隐随机调度中的应用[j]. 电力自动化设备, 2014, 34(3): 125-131
