Identification of Different Stages of Melting in Electric Arc Furnace Using Transformer Tank Vibration Analysis

Accurate identification of melting stages is crucial for optimizing the furnace process control system. This article aimed to develop a new method for accurately identifying melting stages based on vibration signals from the transformer tank. Vibration signals are collected at different stages of the melting process. Using frequency domain analysis, the key features of each stage are extracted from the vibration velocity signal. A novel algorithm is developed to accurately classify melting stages based on the extracted features containing vibration trends and top ten charts. The method is validated through two case studies and shows strong correlation with the total harmonic distortion trend. Compared to conventional methods, the proposed approach provides clearer separation between melting stages and offers improved reliability for real-time monitoring and control, leading to enhanced product quality, and energy efficiency.