A Deep Neural Network Approach for Optimizing Economic Load Dispatch in Power Systems

To ensure minimization of power losses as well as economic feasibility of electrical power generation, economic load dispatch happens to be one of the most challenging optimization problems which is faced in electrical engineering. With the advent of distributed power systems, an interconnection of power systems generating from different sources have come into consideration. However, all sources do not operate in the same manner and hence the generation cost for different sources varies significantly. Economic Load Dispatch (ELD) can be defined as a technique to schedule the power generator outputs with respect to the load demands, and to operate the power system in the most economical way. This paper presents a neural network model for implementing economic load dispatch for a three as well as six generation system. The load is also varied for both the 3 and 6 generations systems. The results clearly indicate that the cost of generation increases with the increase in load, which is also intuitive and hence gets tested. Thus the proposed model can be used to implement an optimized economic load dispatch mechanism for multi-unit power systems which is a typical characteristic of distributed power systems.