Economic Load Distribution Among Generating Plant Using Software Based on Articial Neural Network (ANN)

This paper presents an application system that provided the best load distribution for optimal power flow (OPF) with minimal fuel cost using artificial neural network (ANN). The idea was informed by the application of the principle of equal incremental cost rate for all the generators. The study also demonstrated a new computational model for the exponential function e which was introduced to the Tanh transfer function. A new version of Tanh function was therefore developed called the Tansigmod. It increased the speed of training of the feed forward network by a gain factor of about 3 units. To test the effectiveness of the proposed system, the case of a five bus network for TransAmadi gas turbine power station in Port Harcourt, Nigeria was demonstrated. Result obtained from the system show for example cost saving or Netsave monthly of about 108,520 Naira for a service load of 60MW supply. This benefit was derived from the system due to optimal distribution of load as against equal distribution of load. The application will assist the operators in the gas turbine power stations with the task of planning power generation economically.