An Analysis Method of Feeder Available Capacity Based on Data Mining

Feeder is one of the most important equipment in distribution network, and it is an important mean to evaluate the power supply capability of distribution network. In this paper, the calculation method of the feeder available capacity is obtained by introducing two parameters, the load simultaneous factor and the demand factor, to evaluate the power supply capability. Firstly, the load simultaneous factor of feeder group is predicted by means of clustering analysis and neural network prediction. Then, according to the actual access capacity, loads in feeder group are divided into saturated and unsaturated loads, and the load needful coefficient can be obtained by combing prediction method of grey forecasting and artificial neural network prediction approach. Finally, the feeder available capacity can be calculated by putting the predicted two parameters into proposed model. The calculation results show that proposed method has certain prediction ability, which makes full use of the feeder load flow and considers the reliability of distribution network operation, and can provide meaningful data support for the power supply company.