Condition Monitoring of Drive Trains by Data Fusion of Acoustic Emission and Vibration Sensors

Abstract

Early damage detection and classification by condition monitoring systems is crucial to enable predictive maintenance of manufacturing systems and industrial facilities. The data analysis can be improved by applying machine learning algorithms and fusion of data from heterogeneous sensors. The paper (see <REFERENCE>) presents an approach for a step-wise integration of classifications gained from vibration and acoustic emission sensors, in order to combine the information from signals acquired in the low and high frequency range. A test rig comprising a drive train and bearings with small artificial damages is used for acquisition of experimental data. The results indicate that an improvement of damage classification can be obtained using the proposed algorithm of combining classifiers for vibrations and acoustic emission. This dataset contains vibration data as well as acoustic emission data recorded on a rotating drive train. For vibration measurements, the setup is instrumented with ICP accelerometers (PCB 607A11, 100 mV/g) at the bearing holder frame of the motor and the bearing holder frame of the shaft. Both vibrational sensors are digitized by a DT 9837A USB-DAQ from (Measurement Computing GmbH). This 4-channel signal analyzer is equipped with a 24-bit ADC, that supports sample rates of up to 100 kSPS and is well suited for the vibration measurements. Acoustic emission signals (AE) are acquired by a piezo transducer (Vallen VS30V, 20 kHz – 80 kHz) connected to an AEP5 pre-amplifier for signal conditioning. These signals are digitized with an USB oscilloscope (PicoScope 2204A). This entry-level 2-channel oscilloscope features an 8-bit ADC with sample rates of up to 100 MSPS with an analog bandwidth of 10 MHz.