基于泡沫大小动态分布的浮选生产过程加药量健康状态分析

针对矿物浮选过程泡沫大小分布随着药剂量的改变而动态变化的特点,提出一种基于泡沫大小动态分布特征的具有自学习功能的浮选生产过程加药量健康状态统计模式识别方法.首先,通过泡沫图像分割、气泡尺寸分布核密度估计获得浮选气泡大小的概率密度分布函数,采用无监督的最远邻聚类方法获得典型药剂量添加状态下的气泡尺寸统计分布特征集；然后,采用简单的贝叶斯推理方法获得测试时间段对应的药剂添加健康状态分析识别结果,并根据浮选生产工况状态的动态变化对各典型药剂状态下的气泡大小统计分布特征集进行在线学习修正.实验结果表明,所提出方法能实时获取泡沫尺寸分布的动态变化,实现浮选药剂操作健康状态的自动识别与评价,为进一步实现浮选生产过程的加药量优化控制奠定了基础.

Dynamic bubble-size-distribution-based health status analysis of reagent-addition in froth flotation process

Since the statistical distribution of bubble sizes varies with the dynamic change of the reagent operation in froth flotation process, a statistical pattern recognition method for the health condition analysis of the reagent-addition is presented based on the adaptive learning of the dynamic distribution features of the froth bubble sizes. After the segmentation of the bubble image and the kernel density estimation of the statistical bubble size distribution, the statistical feature sets of the bubble size distribution under typical operating conditions are first leaned by the unsupervised farthest neighbor clustering; the health status of the reagent operation in the flotation process during the period of testing is subsequently inferred by Bayesian inference. Furthermore, the bubble size distribution feature sets under typical dosage-addition conditions are revised online in accordance with the drift of the operation conditions. The experimental results demonstrate that this method is capable of capturing the dynamic variation of the statistical distribution of the bubble sizes and effectively achieving the accurate recognition results of the health conditions of the reagent-addition, which lays a foundation for the realization of the optimal control of reagent-addition in the flotation process operation.