Machine Learning-Based Intelligent Pump Status Monitoring

This paper introduces a pump intelligent condition monitoring system using Raspberry Pi as the controller, integrating data acquisition, processing, and fault diagnosis. The ADXL345 sensor collects vibration signals, and a machine learning algorithm classifies and predicts pre-processed data. Optimization through experiments achieves an error rate of 0.034 and accuracy of 0.974. The system accurately classifies pump states and predicts failures.