British Journal of Mathematics & Computer Science

Abstract:

This paper used a modified Neural Network algorithm for green load balancing. The algorithm applied to electricity distribution problem of Jordan. Neural network learns to predict the demand of electricity in Jordan and allocates electricity according to that prediction. Back-propagation learning algorithm is used to train neural network. Training process depends on Electricity Distribution Parameters, which are: Sector Type, and Weather Type and Daytime. These parameters are selected with respect to Jordan nature. After training, the algorithm always maintains the active servers according to current values of these parameters. As a result, energy consumption will be low and Green computing is achieved.