Dynamic Calculation of Conductor Temperature of Single-cable Using BP Neural Network

It is necessary to accurately calculate the conductor temperature of operating cable for maximizing the use of cable capacity.Therefore,we established a BP neural network model for dynamic calculation of cable conductor temperature by using operating test data of existing 110 kV XLPE single-cable.The model was based on the parameters of actual surface temperature and current,considering the effect of environment.In order to verify the accuracy of BP neural network model,we designed temperature rising test for step current of 110 kV XLPE single-core cable,which was buried in soil and laid in the air.Then calculation of conductor temperature from BP neural network model were compared with measured data of conductor temperature,and the errors between them were analyzed,in the meantime,calculating results of the BP neural network model,Laplace algorithm,ANSYS algorithm and IEC-60287 algorithm were compared.The results show that the dynamic calculation of conductor temperature of BP neural network model can be used in calculating cable conductor temperature based on its surface temperature and actual current,without taking its physical parameters into account,moreover,it also can give a reference on accurately monitoring the state of operating cable.