An Optimized Back Propagation Neural Network Based on Genetic Algorithm for Line Loss Calculation in Low-voltage Distribution Grid

A calculation model of the theoretical line loss rate of smart distribution network in the low-voltage substation area was devised. Based on the back propagation neural network, the model was optimized by the genetic algorithm (GA-BP model). The model took the active and reactive power of each power supply area as the input and the theoretical line loss rate as the output. The results showed that the genetic algorithm had achieved the best fitness after 92 iterations, and the R-squared of all training data was R=0.99916; after calculated the test samples, the maximum absolute error was 0.14% of GA-BP model, which was 0.61% less than that of the BP model; the mean absolute error was 0.055%, which was 0.273% less than that of BP model. In general, the GA-BP model has a better non-linear fitting ability and higher accuracy of calculating the theoretical line loss rate in the low-voltage distribution area.