Research on transformer fault diagnosis based on GWO-RF algorithm

Efficient fault diagnosis of power transformer can effectively ensure the safe and stable operation system. Considering that effect traditional Random Forest (RF) is seriously affected by initial parameters, this paper proposes a RF algorithm based on Grey Wolf Optimization (GWO-RF) to improve accuracy identification. The uses GWO optimize total number decision trees depth maximum tree, balancing model with ability generalize. In paper, Dissolved Gas in oil taken as characteristic quantity, 335 sets data are used form set. Three different verified GWO-RF, IEC three ratio method. experimental results show effectiveness GWO-RF