Research on application of artificial neural network in the network loss calculation of power system

Power losses in an electrical supply grid have significant economic implications for the performance of power system. In a vertically integrated power structure, the system operator can minimize the total network losses by adjusting active and reactive outputs of all generators connected the system according to the results of optimal power system. In a competitive environment, restructuring of the power system leads to unbundling of generation, transmission and distribution into various autonomous business units. So sharing of the network losses is no longer a simple job and its allocation among end-users is a challenging job. Based on the technology of artificial neural network (ANN), this paper presents a new approach for network lose calculation of power system. This method can realize complicated nonlinear mapping relations after training, which needs not write out the complex relationship between input and output, and so it reduces the amount of calculation and speeds up the calculation greatly. An calculation example showed its feasibility.