Neural network based distributed generation allocation for minimizing voltage fluctuation due to uncertainty of the output power

The problem of distributed generation (DG) allocation and sizing is of great importance, since improper integration of DG units cause to take a bad turn in terms of power quality and system efficiency at high penetration levels. In this reason, allocation of DG is not a trivial optimization problem. The reactive and active power fluctuation of DG will lead to voltage fluctuation, especially for wind or photovoltaic power generators. Their output powers are more unpredictable due to the intermittent wind speed and irradiation. In this paper, the effects of reactive and active powers of DG on voltage profile are analyzed by including their output power fluctuation and an artificial neural network (ANN) based decision support system are developed to be used in management and planning of DG integration. The proposed system can be used to determine suitable bus to reduce the voltage fluctuation of critical buses. The simulation results presented shows the effectiveness of the method.