基于灰度图像特征的电选粉煤灰烧失量预测  

Prediction for the Loss on Ignition of Elective Fly Ash Based on Gray Image Features

作　　者：陈师杰 李海生[1,2] 陈英华 温晓龙[1] 郑诚[1] 王文平 CHEN Shijie;LI Haisheng;CHEN Yinghua;WEN Xiaolong;ZHENG Cheng;WANG Wenping(School of Chemical Engineering and Technology,China University of Mining and Technology,Xuzhou,Jiangsu 221116,China;Key Laboratory of Coal Processing and Efficient Utilization of Ministry of Education,Xuzhou,Jiangsu 221116,China)

机构地区：[1]中国矿业大学化工学院,江苏徐州市221116 [2]煤炭加工与高效洁净利用教育部重点实验室,江苏徐州市221116

出　　处：《矿业研究与开发》2020年第2期145-149,共5页Mining Research and Development

基　　金：国家自然科学基金项目(51674259)

摘　　要：针对传统粉煤灰烧失量检测耗资大,耗时长等问题,提出了粉煤灰烧失量图像识别快速检测的方法,运用工业摄像头获取不同梯度烧失量的粉煤灰图片,通过MATLAB处理后来获取粉煤灰的灰度图像特征参数,根据灰度特征参数,分别利用BP神经网络、SVM神经网络、ELM神经网络建立粉煤灰烧失量的预测模型,比较不同神经网络对粉煤灰烧失量的预测效果。结果表明:通过ELM神经网络建立的粉煤灰烧失量模型对粉煤灰烧失量的预测效果最好,该模型稳定性好,预测准确度高,可用于工业生产中的粉煤灰烧失量快速检测。The traditional loss on ignition test of fly ash will cost a lot of money and time.For this problem,The scheme of image recognition rapid test for loss on ignition of fly ash was proposed.The images of fly ash with different gradient loss on ignition were obtained by industrial camera.The gray image feature parameters of fly ash were obtained by MATLAB.According to gray feature parameters,BP neural network,SVM neural network and ELM neural network were used to establish the prediction models for the loss on ignition of fly ash.And the prediction effects of different neural networks on the loss of fly ash were compared.The results showed that the model established by the ELM neural network had the best predictive effect,with a good stability and a high prediction accuracy,which can be used for rapid test of loss on ignition in industrial production.

关 键 词：粉煤灰 摩擦电选 烧失量 图像识别 神经网络 

分 类 号：TD94[矿业工程—选矿]