基于纹理单元分布的硫浮选过程病态工况检测

凌弈秋; 阳春华; 何明芳; 桂卫华

基于纹理单元分布的硫浮选过程病态工况检测

引用本文：凌弈秋，阳春华，何明芳，桂卫华.基于纹理单元分布的硫浮选过程病态工况检测[J].计算技术与自动化,2013,(1):28-31

摘要点击次数: 2335

全文下载次数: 96

作者	单位
凌弈秋，阳春华，何明芳，桂卫华	(中南大学信息科学与工程学院,湖南长沙410083)

中文摘要:针对硫浮选过程难以及时发现病态工况的问题，提出一种基于泡沫图像纹理单元分布函数的硫浮选过程病态工况检测方法。通过计算纹理单元值提取泡沫图像的纹理特征，避免传统灰度共生矩难以准确描述纹理特征的问题，设计核函数估计算法逼近泡沫纹理单元概率密度函数(PDF)，使得纹理单元分布函数转化成一组动态权系数。建立基于纹理单元分布动态权系数的主元模型，并基于该模型的T2统计量得到检测阈值，从而实现硫浮选过程病态工况的及时检测。湿法炼锌直接浸出硫浮选工段现场运行结果表明此算法的有效性。

中文关键词:纹理单元核密度估计主元模型

Fault Condition Detection for Sulfur Flotation Process Based on Texture Unit Distribution

Abstract:Considering the difficulty to detect the fault condition of sulfur flotation in real-time, a new fault detect method based on probability density function (PDF) of froth texture unit is proposed. The froth texture unit is used to describe the froth texture features to avoid that the gray level dependence matrix can not describe the froth texture features accurately. By means of the designed kernel method，the PDF is transformed into dynamic weight coefficients, based on which a principal component analysis (PCA) model is established. And the T2 statistical quantities based on PCA model can be used as a threshold criterion to realize the fault condition detection in sulfur flotation process. Simulation results show the effectiveness of this algorithm by using actual running data from the sulfur flotation process of zinc direct leaching hydrometallurgy.

keywords:texture units kernel estimator principal component analysis model

查看全文 查看/发表评论 下载pdf阅读器