Energy management in microgrids using IoT considering uncertainties of renewable energy sources and electric demands: GBDT-JS approach

The energy management issue of microgrids typically adopts demand response programs and reconfiguration of distribution networks to improve the technical and financial characteristics of microgrids. This manuscript proposes an energy management optimization in micro grids using IoT by applying the GBDT-JS technique to account for the uncertainty introduced by renewable energy sources and electric demands. The proposed technique is a combination of gradient boosting decision tree algorithm (GBDT) and jellyfish search (JS); hence, it is called GBDT-JS technique. The proposed method is to include maximizing the benefits of microgrids and reducing power fluctuations to the main grid. This proposal presents a GBDT-JS to develop the suitable adaptive day-ahead real-time energy dispatch under the occurrence of operating uncertainty based on the energy management system (EMS). Energy dispatch is achieved to satisfy the many operating criteria based on optimum solutions which are further implemented, for example minimum power fluctuation and operating cost. The main aim of the works contributions is shortened as below: (1) The optimum GBDT system is decided day-ahead depending on the predictive microgrid system levels consider a many operating objects. (2) JS technique is improved a day-to-day basis, adopted to execute the energy dispatch, and directly present in the uncertainty system. The GBDT-JS technique is done on MATLAB, and their performance is compared to different existing techniques, such as slime mould optimization algorithm, side-blotched lizard algorithm, and jellyfish search (JS).