首页|激光诱导击穿光谱技术预测煤的黏温特性

激光诱导击穿光谱技术预测煤的黏温特性

Prediction of Viscosity-Temperature Characteristics of Coal by Laser Induced Breakdown Spectroscopy

扫码查看
快速准确测定煤的黏温特性对气化炉操作具有重要意义,通常采取将煤烧灰测定其临界黏度温度的方法来确定煤的黏温特性.激光诱导击穿光谱(LIBS)与化学计量方法结合在煤炭分析方面有许多独特优势.收集41个代表性煤样,基于DL/T 660-2007标准方法,测定相应煤灰的临界黏度温度,将其与煤样的LIBS谱图关联,利用化学计量学方法建立预测煤灰临界黏度温度的LIBS分析模型.模型交互验证结果表明,LIBS方法与实验室标准方法之间具有良好的一致性,基于偏最小二乘回归(PLSR)、人工神经网络(ANN)和支持向量机回归(SVR)建立的分析模型对临界黏度温度的预测标准偏差分别为63、57、56 ℃,单个煤样的测试时间不超过5min,满足工业现场快速分析的需求.
To rapidly and accurately determine the viscosity-temperature characteristics of coal is of great significance for gasifier operation.This is usually achieved by the method of measuring the temperature of critical viscosity of coal ash.The combination of laser induced breakdown spectroscopy(LIBS)and chemometrics methods has many unique advantages in coal analysis.The study collects 41 representative coal samples,based on which the corresponding temperature of critical viscosity of coal ash has been determined by DL/T 660-2007 standard method.Based on correlation between the laser induced breakdown spectra of coal sample and the temperature of critical viscosity,three LIBS calibration models for predicting the temperature of critical viscosity are built by chemometrics methods.The cross validation results of the models show good consistency between LIBS method and standard laboratory method.The standard error in prediction of temperature of critical viscosity from the calibration models established by partial least squares regression(PLSR),artificial neural network(ANN),and support vector machine regression(SVR)is 63,57,and 56 ℃,respectively,and the measurement time for each coal sample is less than 5 minutes,thus meeting the needs of rapid onsite analysis for industry.

laser induced breakdown spectroscopycoalchemometricstemperature of critical viscosityviscosity-temperature characteristics

许育鹏、汪天也、王树青

展开 >

中石化石油化工科学研究院有限公司,北京 100083

激光诱导击穿光谱 化学计量学 临界黏度温度 黏温特性

中国石化合同项目

321137

2024

石油学报(石油加工)
中国石油学会

石油学报(石油加工)

CSTPCD北大核心
影响因子:0.764
ISSN:1001-8719
年,卷(期):2024.40(3)
  • 19