Hybrid PSO-BP Based Probabilistic Neural Network for Power Transformer Fault Diagnosis

Diagnosis of power transformer abnormality is very important for power system reliability. This paper presents a novel approach for power transformer fault diagnosis based on probabilistic neural network and dissolved gas-in-oil analysis (DGA) technique. A new hybrid evolutionary algorithm combining particle swarm optimization (PSO) algorithm and back-propagation (BP) algorithm, referred to as HPSO-BP algorithm, is proposed to select optimal value of PAW parameter. The HPSO-BP algorithm is developed in such a way that PSO algorithm is used to do a global search to give a good direction to the global optimal region, and then BP algorithm is used as a fine tuning to determine the optimal solution at the final. The experimental results show that the proposed approach has a better ability in terms of diagnosis accuracy and computational efficiency compared with a number of popular fault diagnosis techniques.