Application of neural network and weighted improved PSO for uncertainty modeling and optimal allocating of renewable energies along with battery energy storage

Abstract

The power uncertainty of renewable energy encounters their optimal allocation with obstacle in an on-grid microgrids. Hence, power prediction is a logical tactic. Besides, optimal performance of battery energy storage for compensating electrical loads is another strategy. However, these two solutions must meet the economic expectations as well as reliability of customers. In this paper, generating section includes wind, solar and wave energies in which wind speed and solar irradiance are uncertain parameters and information of Hormoz Island, Iran is the input of problem. Hence, artificial neural network (DANN) is trained by three adaptive techniques to minimize the prediction error dynamically. Then, eco-statistic objective function, reliability criterion and efficient battery strategy are modeled thoroughly. After introducing five feasible scenarios, a heuristic algorithm which is weighted improved PSO, searches the global answers of optimization to test the efficiency of hybrid architecture. Moreover, forecasting results are compared with five adaptive neural network based fuzzy inference system (ANFIS) and one conventional Levenberg Marquardt (LM) based ANN for addressing the functionality of proposed DANN against the uncertainty of renewables. Consequently, simulation results of optimization are tested with heuristic algorithms to demonstrate the optimal answers of problem.