Condition Monitoring of Air Compressor in Steel Industry using ANFIS

As large rotating machines are increasingly employed in continuous operations at high speeds and with heavy loads, vibration behavior of rotating systems is emerging as more complex phenomenon. Monitoring vibration behaviour of large rotating machinery is an effective way to reduce losses and enhance safety, reliability, availability and durability in manufacturing processes. This research focuses on condition monitoring of one of the vital and the most critical machine Air Compressor of steel industry. It considers vibration levels of the machinery based on ISO limit of vibration severity using Adoptive neuro fuzzy inference system (ANFIS). Two different data schemes were formulated based on preliminary experimentation on Sugeno type 3 inputs (vHm, vVm & vAm) and 1 output (i. e., Condition) ANFIS model. The performance criterion of the ANFIS classifier was evaluated using confusion matrix. The total classification accuracy of 95% obtained proves the validation of the Air Compressor model. ANFIS can also be extended to condition monitoring of various rotating machinery.