A mill vibration and vibro-acoustic feature extraction method based on multi-source information

Abstract

The invention discloses a mill vibration and vibration-acoustic feature extraction method based on multi-source information, which includes the steps: A, multi-source signal acquisition, which obtains the signal representing the load state of the mill through a data acquisition system, and the collected The signal is a multi-source time-domain signal, which is denoted as S, and the multi-source time-domain signal S with a sampling number of N is denoted as B. Shallow feature information selection, which is based on non-stationary nonlinear signal adaptive decomposition and analysis technology Realize the selection of shallow feature information of mill vibration and vibro-acoustic; C: Deep feature knowledge extraction, which realizes the extraction of mill vibration / vibro-acoustic multi-scale deep feature knowledge based on optimized coupling deep neural network feature extraction and mutual information feature measurement algorithm . The method of the invention can provide support for the construction of mill load state identification and mill load parameter soft-sensing model.

Description

Technical field [0001] The invention relates to the technical field of mill grinding process, in particular to a method for extracting characteristics of mill vibration and vibration and sound based on multi-source information. Background technique [0002] Grinding is the "bottleneck" operation of the beneficiation process. Its function is to grind the crushed ore into a slurry of qualified particle size through rotating mechanical equipment (ball mill) to provide raw materials for subsequent processes. The load of the mill determines the production efficiency and quality indicators of the grinding process and even the beneficiation process. Overloading of the mill will cause "mill spit out", "greater granularity at the exit of the mill", "grind blocking", and even stop production. The underload of the mill will cause the mill to "empty hit", resulting in energy waste, increased steel consumption, and even damage to the grinding equipment. It is generally believed that the rea...

AI Technical Summary

Problems solved by technology

However, domain experts cannot directly perceive signals such as mill cylinder vibration and bearing vibration, and there are large differences in the experience of different domain experts

Therefore, experts in the field cannot use the highly sensitive cylinder vibration, the bearing vibration with weak random interference, and the mill current with extreme point characteristics by virtue of the auditory perception mechanism.

Images