Neural Network Based on Artificial Intelligence Solution for a Fast Economic Load Dispatch Using the Hybrid Lagrangian Method

Abstract

This paper comes up with a new approach in order to solve the Economic Load Dispatch problem by using a hybrid technique. This technique combines the artificial neural network with the Lagrangian multiplier method. The adaptive behavior of the controller is obtained through the action of the neural network on the Lagrangian multipliers’ parameters. The suggested technique has been implemented on 10 test-system units with different constraints like the spinning reserve, the generator capacity limits, the start-up cost, the prohibited zones, the minimum up and the minimum down time constraints and power balance. The results obtained from the proposed approach are compared with the ones of the conventional Lagrangian multiplier method. They prove the efficiency, the fast convergence with less computational time and, accordingly, the proficiency of the suggested approach.