Power systems planning, operation and control are tightly related to load characteristics and voltage magnitudes. Voltage stability is directly dependent on load behavior with voltage and frequency variations. This article presents a methodology for modeling an Egyptian primary distribution system load dynamics using artificial neural networks (ANN). Real field measurements of load responses (voltage/time and power/time) have been recorded during system transients by means of a digital fault recording (DFR) device. These measurements are used to train and test three proposed ANN-based load models for the chosen distribution system. The ANN simulation results are illustrated and compared with those obtained from the DFR. The results show that ANN-based load models can emulate load dynamics accurately.
Available at: http://works.bepress.com/almoataz_abdelaziz/25/