计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (1): 228-230.

• 工程与应用 • 上一篇    下一篇

基于机器视觉的矿物浮选pH软测量方法

阳春华,任会峰,桂卫华,周开军   

  1. 中南大学 信息科学与工程学院,长沙 410083
  • 收稿日期:2010-03-23 修回日期:2010-05-23 出版日期:2011-01-01 发布日期:2011-01-01
  • 通讯作者: 阳春华

Machine-vision-based soft sensor of pH for flotation process

YANG Chunhua,REN Huifeng,GUI Weihua,ZHOU Kaijun
  

  1. College of Information Science and Engineering,Central South University,Changsha 410083,China
  • Received:2010-03-23 Revised:2010-05-23 Online:2011-01-01 Published:2011-01-01
  • Contact: YANG Chunhua

摘要: 针对泡沫浮选过程中人工检测矿浆pH值严重滞后以及pH检测仪电极容易积垢导致测量不准等问题,提出基于机器视觉的浮选矿浆pH软测量方法。确定了与矿浆pH值最相关的图像特征泡沫颜色、尺寸和流速;采用减法聚类确定模糊系统的初始结构,并选择变尺度分级混沌方法优化隶属函数和输出权值,最终建立模糊神经网络pH软测量模型。工业实践证明应用该模型在线检测浮选矿浆pH值的可行性。

Abstract: Considering the serious time-delay and rough detection of pH by operators,a soft sensor of pH based on machine vision is presented.Froth color,size and velocity are selected as the most relative features with pH.Furthermore,the initial structure of fuzzy system is given using subtractive clustering.The correlation of image features and pH is described using membership function.Parameters are optimized by using variable metric hierarchy chaos algorithm.The soft sensor model based on fuzzy neural network is set up.The results show that the proposed model performs well detection precision online.