The line loss is a significant contributor to measuring the energy conservation, emission reduction, and operational efficiency of the distribution system. As China's “dual-carbon” goal continues to advance, large amounts of stochastic and intermittent renewable energy access to the distribution network, giving a new challenge to the line loss computation of distribution networks. Through a BP neural model application, the distribution network's line loss value in a certain operation state is calculated by taking the load value at every node of the distribution network and the 0–1 discrete value for every line in the distribution network whether it is put into operation or not as input variables. The analysis of the arithmetic example based on the IEEE14 node system shows that the method presented enables accurate and efficient calculation of line loss values for power systems under specific operating conditions.