Distribution Network Fault Diagnosis Based on Rough Set theory and BP Algorithm
Kan Xu · Journal of Electric Power · 2014
Combined rough set theory with BP neural network, information systems decision table of rough set as the main tool, rough set attribute reduction algorithm base on genetic algorithms to reduce distribution network raw data. Then use BP network for the most simple set of rules for learning and training, Finally,use improved particle swarm optimization to optimize the weights and thresholds of BP network, form a model structure based on rough set and BP networks with improved particle swarm optimization. The experimental results show that through rough set attribute reduction, delete a lot of unnecessary redundancy data, improve the speed and accuracy of diagnostic. Meanwhile,for the distribution network failure exists malfunction,tripping,etc.,the algorithm has higher generalization ability and better fault tolerance.