An Efficient Feed-Forward Neural Network for Economic Load Dispatch in Power Systems

This paper provides an attempt to explore the ability of Feed - forward Neural Network to solve one of the complex problems of dispatching the electric load amongst the available generators. Economic Load Dispatch is basically a problem of optimizing a non-linear cost curve of generator. Artificial Neural Networks is a very much useful tool when it comes to the non-linear optimization. Two constraints namely, active power limits of each generators and generator-load balance have been observed while preparing the model. The prepared model has been tested on 9-bus system for variety of loading conditions. The results obtained were compared with traditional numerical approach to solve the economic load dispatch problem. The results are plotted in terms of error, which compares the output of proposed method and traditional method. Simulation results illustrate the efficiency of proposed approach.