Intelligent framework and techniques for power transformer insulation diagnosis

This paper presents an on-going project of developing a suite of intelligent diagnostic tools for evaluating power transformer insulation systems. Our ultimate goal is to exploit techniques for automatically processing the measurement data, extracting meaningful information and subsequently transforming the information into domain-specific knowledge which can be utilized for indicating power transformer fault conditions and their serviceability. In this paper, an intelligent diagnostic framework targeting power transformer insulation system evaluation is proposed. Several pattern classification algorithms for power transformer faults recognition are shown. Simulation results based on the on-site experimental data are presented.