Design of an Adaptive Optimizer for Economic Load Dispatch in Power Systems

Abstract

The rapid rise in global power consumption necessitates cost-effective power generation. Economic Load Dispatch (ELD) plays a crucial role by optimizing the load distribution across power sources to minimize generation and transmission costs. This study presents the steepest descent-based neural network model to achieve economic load dispatch by optimizing power generation within system constraints. A scaled conjugate gradient approach is applied to train the neural network, with simulations conducted on systems of 3 and 6 generating units. Results indicate that generation costs increase with load, aligning with theoretical expectations and highlighting the model's potential for reducing costs in power systems.