Adaptive Neuro-fuzzy Inference System Structure Research Articles

A HYBRID APPROACH TO HYDRAULIC VANE PUMP CONDITION MONITORING AND FAULT DETECTION

Health monitoring of hydraulic vane pumps is currently achieved through oil analysis. Monitoring pump operating parameters to detect faults on-line could improve condition-based maintenance techniques. This research developed an on-line hydraulic vane pump fault detection system. This fault detection system decomposed vertical pump vibration signals using wavelet packet analysis. Packets containing signal features distinguishing normal and failed pump operation were entered into an adaptive neuro-fuzzy inference system (ANFIS) for pump health classification. Two particular vertical vibration signal packets (7,11 and 7,60 from the Daubechies 3 packet analysis) were found most effective in classifying pump condition when entered into the ANFIS structure. Pump health classification using these inputs was 90% accurate with a missed alarm rate of 3% and a false alarm rate of 6%.