Neural network approach for estimation of load composition

A neural network methodology to solve the problem of estimation of modern electrical load distribution in typical commercial and residential areas is proposed in this paper. The inputs for the neural network are harmonic characteristics of each type of typical loads and possible combinations of these loads. The output is the estimation of load composition. The multi-layer feed-forward back-propagation neural network and Elman neural network are used in the paper to calculate the load distribution. A case study of a Manhattan area and two practical tests are presented to demonstrate the feasibility of this approach. The new method is useful for electrical load monitoring and harmonic reliability assessment in the new utility environment.