Analysis of Defects' Impact on the External Magnetic Field of the Turbogenerator

This paper presents an approach to monitoring the technical state of turbogenerator based on the control of an external magnetic field. The study had been performed for the synchronous turbogenerator T3F-160 and includes computer modeling and using neural networks. The effects of defects in form of winding current asymmetry and tooth fracture in the magnetic core, are examined. Simulation results demonstrate the feasibility of defect identification through deviations in the magnetic flux density distribution. Based on the obtained data, a neural network was developed to classify machine states with defect-specific accuracy. The proposed method enhances the reliability of diagnostics and fault prediction, enabling integration into existing monitoring systems.