Technical losses computation for short-term predictive management enhancement of grid-connected distributed generations

the integration of distributed generations in medium voltage feeders is conditioned by multiple rules, especially, by those related to power flow management through the network and the efficient handling of renewable energies intermittency. For that purpose, we proposed in this work an active management algorithm to predict the need in terms of the power to be injected by a High Voltage/Medium Voltage substation for every single feeder issued from this substation. We apply it on a medium voltage feeder, considering that this feeder contains a photovoltaic installation and a storage system. The developed algorithm will allow us to underpin the forecast accuracy results through the adoption of many approaches. Those approaches aim to adjust the load demand forecast, ensure a reliable photovoltaic power production prediction, estimate the technical losses of the system and manage economically and optimally the energy flow in the battery storage bank. The present study will permit, from the one hand, to minimize energy losses in a grid connected distributed generations. The latter can be realized by managing in advance the energy flow between the different medium voltage feeders and substations for each region. Then, predict the whole energy production from conventional sources at the National Dispatching level. From the other hand, it will allow us satisfy load demand while avoiding peaks and rising the electrical devices lifetime by optimizing the number of operations on the network and forecasting the different suitable regulations. To succeed those objectives, we tried to find the best way to minimize the prediction error of our model. We exclusively adopt a particular approach to define each key performance indicator. We cite mainly the estimation of the technical losses in distribution and production segments, separately, the forecast of the load demand by considering the impact of weather conditions, the evaluation of the impact of cloud motion on PV panels and finally integrating those parameters into a MPC to enhance the accuracy of the transformer output prediction.