Material level prediction method for ball mill

Abstract

The invention discloses a material level prediction method for a ball mill. The method comprises the following steps that: collecting a vibration signal on the cylinder wall or the bearing of the ballmill and the sound vibration signals of microphones around the ball mill; inputting the vibration signal and a sound vibration signal into a data collection card through a conditioning circuit and ananalog-to-digital conversion circuit; through an MFCC (Mel-Frequency Cepstral Coefficients, independently calculating the feature parameters of the vibration signal and the sound vibration signal inthe data collection card; re-combining the vibration signal with the sound vibration signal; collecting the re-recombined signal by the data collection card; and finally, predicting the material levelof a corresponding signal for data in the data collection card through a TBM (Tactical Ballistic Missile) model established by historical data. By use of the method, the material level of the ball mill can be effectively predicted, and the method has the advantages of high material level prediction accuracy, small fluctuation and good practical value and popularization prospect.

Description

technical field [0001] The invention relates to the field of data detection, in particular to a ball mill material level prediction method. Background technique [0002] Roller ball mill is a basic milling equipment. The material level inside the ball mill drum is the key factor affecting the energy consumption, production efficiency and production safety of the ball mill. Accurate prediction makes the ball mill system difficult to operate economically and prone to safety accidents. Therefore, it is of great significance to effectively predict the material level of the ball mill for the efficient, safe and stable operation of the ball mill system. [0003] At present, the ball mill material level measurement method based on soft sensor mainly predicts the material level through a single signal source such as piezoelectric signal, differential pressure signal, vibroacoustic signal, vibration signal or power signal. From trial and error, we found that a single signal source...

AI Technical Summary

Problems solved by technology

From trial and error, we found that a single signal source has certain limitations. For example: the continuous increase in the use time of the piezoelectric signal, the piezoelectric strain sensor is prone to zero drift, resulting in a decrease in measurement accuracy; the differential pressure signal requires more auxiliary variables. , each variable value is difficult to obtain accurately, there is redundancy and coupling between variables, and the system complexity is high, which is inconvenient to maintain; the vibro-acoustic signal is eas

Images