ABSTRACT
These good results encouraged the adoption of the described neural system for the actual use, which is still in use at the present time.
TABLE VI.5.8 Errors Between the Differences of Neural Interpolation and the Linearized Load, and the Recorded Load and the Same Linearized One
Several reasons can be given for the suitability of adaptive algorithms for load forecasting problems:
1. ANNs are a mature technique to be used in load forecasting. Several utilities in the world have already adopted ANN models in daily use, with reported advantages in prediction accuracy. The most adopted practice is the multi-layer feedforward scheme.
