Short term load forecasting: A dynamic neural network based genetic algorithm optimization
MetadataShow full item record
CitationWang, Y., Ojleska, V. M., Jing, Y., Gugulovska, T. D. K., & Dimirovski, G. M. (2010). Short term load forecasting: A dynamic neural network based genetic algorithm optimization. In 2010 14th International Power Electronics and Motion Control Conference (EPE/PEMC) (pp. T6-157-T6-161). Piscataway, NJ: IEEE. https://dx.doi.org/10.1109/EPEPEMC.2010.5606508
The short term load forecasting plays a significant role in the management of power system supply for countries and regions, in particular in cases of insufficient electric energy for increased needs. A back-propagation artificial neural-network (BP-ANN) genetic algorithm (GA) based optimizing technique for improved accuracy of predictions short term loads is proposed. With GA's optimizing and BP-ANN's dynamic capacity, the weighted GA optimization is realized by selection, crossing and mutation operations. The performance of the proposed technique has been tested using load time-series from a real-world electrical power system. Its prediction has been compared to those of obtained by only backpropagation based neural-network techniques. Simulation results demonstrated that the here proposed technique possesses superior performance thus guarantees better forecasting.