The research presented proposes a new model for predicting electricity demand based on artificial neural networks with the Backpropagation Resilient algorithm. The aim of the research is to evaluate the accuracy of this model compared to a previous model based on artificial neural networks with the traditional Backpropagation algorithm. A field observation was made and 70128 observations of electricity demand were collected. Of these, 61344 were used to train the model and 8784 were used to assess its accuracy. After preprocessing data to correct for missing and atypical data, the results of both models were compared in terms of mean absolute error percentage (MAPE). The Backpropagation Resilient neural network-based model was found to have an ASM of 2.47%, while the traditional Backpropagation-based model obtained an ASM of 2.63%. These results suggest that the Backpropagation Resilient neural network-based model is more accurate than the traditional Backpropagation-based model. In conclusion, the results of this research suggest that the model based on the Backpropagation Resilient neural network is a more accurate option to predict the demand for electrical energy and it is recommended to perform an adequate preprocessing of the data before designing any neural model to obtain better results. Importantly, the research presented must be replicated and validated by other studies before it can be applied in the electric power industry