Study on Pump Fault Diagnosis Based on Rough Sets Theory

In this paper, a rough classifier based on rough sets theory is studied and employed to diagnose and identify five-plunger pump faults. To do so, the spectrum features of vibration signals collected in the flood end of the pump are abstracted as the attributes of the learning samples. Then attribute reduction is carried out to generate the decision rules used to classify technical states of considered object. The diagnostic investigation is done on data from a fivepump in outdoor conditions on a real industrial object. Results show that the new approach can effectively identify different operating states of the pump, which supplies as the basis for the detection and diagnosis of the pump faults.