Abstract: The main objective of this study is to give the pollution issue the priority in terms of cost calculations and optimizations. This paper presents an effective method based on using the neural networks to reduce the impact of air pollution by rearranging the stations within the system by its minimizing the emissions based on a penalty-pollution function, as a mitigation measure which reflect the environmental situations. Therefore satisfy the optimum economic dispatching loads between the stations. The neural network solves multi complex problems of the electrical system adapting with continuous varying of conditions in real time.