The feature extraction method based on the consistency weighted fusion algorithm for ball mill load measurement

Ball mill load is the most important parameter in process monitoring of ball mill, and the direct measurement of ball mill load is difficulty. A novel method of feature extraction based on the consistency weighted fusion algorithm for ball mill load measurement is proposed in the paper. The new method is according to the principle of multi-sensor consistency measuring and the method is suitable for the analysis of ball mill noise signal or vibration signal. At first, the power spectrum of signal is calibrated by several frequency bands, and amplifies the information content of high frequency band. Then the similarity of several frequency band data is obtained by the improved grey incidence analysis algorithm, and the characteristic frequency band which is consisted with ball mill load is gotten. Finally, the feature of ball mill load is calculated by the least square weighted fusion algorithm. Experimental results show that the proposed method is efficient.

[1]  Yong Yan,et al.  A wavelet-based multi-sensor data fusion algorithm , 2003, Proceedings of the 20th IEEE Instrumentation Technology Conference (Cat. No.03CH37412).

[2]  Ren C. Luo,et al.  Dynamic multi-sensor data fusion system for intelligent robots , 1988, IEEE J. Robotics Autom..

[3]  Manus P. Henry,et al.  The fusion of redundant SEVA measurements , 2005, IEEE Transactions on Control Systems Technology.

[4]  C. V. R. Murty,et al.  Experimental analysis of charge dynamics in tumbling mills by vibration signature technique , 2007 .

[5]  Sifeng Liu,et al.  Review of the algorithm models of degrees of grey incidence , 2008, 2008 IEEE International Conference on Fuzzy Systems (IEEE World Congress on Computational Intelligence).