基于FCM-LSSVM算法的球磨机状态预测研究
Research on Ball Mill State Prediction Based on Kalman Filter Improved Least Squares Support Vector Machine(FCM-LSSVM)Algorithm
刘春辉 1盖俊鹏 1胡健 1王迎镇 2张兴帆3
作者信息
- 1. 鞍钢集团关宝山矿业有限公司
- 2. 北京科技大学矿产研究院
- 3. 中国科学院沈阳自动化研究所
- 折叠
摘要
球磨机的稳定运行对选别作业的稳定给料和选别效益提升至关重要,现有的预测方法难以实现磨机状态的快速检测与准确识别.通过卡尔曼滤波改进最小二乘支持向量机方法建立了某铁矿选厂球磨机健康状态模型.根据选厂历史记录数据,采用K-means聚类算法训练得到球磨机的4种健康状态,实现了现场球磨机运行健康状态的实时快速识别.
Abstract
The stable operation of the ball mill is very important for the stable feeding and the improve-ment of the sorting efficiency.The existing mill state prediction method is difficult to realize the rapid detec-tion and accurate identification of mill state.The health state model of a ball mill in an iron ore concentrator is established by using Kalman filter modified least squares support vector machine method.According to the historical data of plant selection,four kinds of health state of ball mill are obtained by K-means cluster-ing algorithm training,and the real-time and rapid recognition of the running health state of field ball mill is realized.
关键词
球磨机/健康状态预测/卡尔曼滤波/最小二乘支持向量机Key words
ball mill/health state prediction/Kalman filter/least squares support vector machine引用本文复制引用
出版年
2024
NETL
NSTL
万方数据