A Neuro-fuzzy Model for Electric Power Line Losses Reduction in Solar Photovoltaic Integrated Sub-transmission Networks

Shortage of sub-transmission system capacities along with the need for reliable power supply have caused increased interest in distributed generation especially using solar photovoltaic source which is considered as clean energy. This paper investigated improvements in losses reduction using solar photovoltaic energy source. A test case sub-transmission network of Volta River Authority/Northern Electricity Distribution Company in the Northern Region of Ghana was used for the modelling and simulations. Power World Simulator Version 19 software was used to model the network. From the Power World simulator software simulations, data on voltage profile and active line losses were extracted and purposely used to train a neural network which output served as input to a fuzzy logic system in a Matlab/Simulink software version 2014b environment. The neural network determined the most sensitive buses whilst fuzzy logic determined the optimal size of solar PV for placement in order to improve the voltage profile and reduce the line losses. Results indicated that, total active and reactive power losses reduced by 25.0% and 14.7% over the base case of no solar PV injection and optimal placement. The voltage profile of the entire length of the network saw an improvement to the value of 1 pu as a result of the solar PV integration using neuro-fuzzy algorithm. Neuro-fuzzy technique is suitable for sub-transmission line losses reduction.