Crompton Greeves Three Phase Pump-Predictive Maintenance Dataset

This dataset contains synthetic data to simulate the continuous operation of the Crompton Greaves CG-IPM-15 three-phase industrial centrifugal pump. The dataset was generated using a hybrid protocol combining GPT-4O-based generative logic and stochastic modeling, overcoming the limitations of field data acquisition. The CG-IPM-15 operates across thermal, electrical, mechanical, magnetic, and hydraulic domains. The synthetic dataset includes 26 sensor streams, representing electrical, mechanical and thermal, and hydraulic behavior. The dataset includes 26 key parameters recorded from a three-phase pump motor system. These parameters are Timestamp, which records the time of each entry; Phase1_Voltage, Phase2_Voltage, and Phase3_Voltage, representing the voltage levels in all three phases; and Phase1_Current, Phase2_Current, and Phase3_Current, which capture the corresponding phase currents. Active power and reactive power measure the real and imaginary power consumed, while power factor indicates the efficiency of power usage. Frequency reflects the operating electrical frequency of the motor. Thermal characteristics are logged using stator temperature, bearing temperature, and ambient temperature. Inlet pressure and outlet pressure track the pressure on both sides of the pump. Flow rate measures the amount of fluid being pumped per unit time. Mechanical vibrations are captured through Vibration X-axis, Vibration Y-axis, and Vibration Z-axis. Motor speed and torque are used to monitor the rotational behavior and mechanical load. Additional health parameters include insulation resistance, oil level, current unbalance, and THD current (total harmonic distortion in current). Bearing FFT harmonics provides frequency-domain insights into bearing health, and failure code denotes categorized fault conditions.