Artificial neural networks for load flow and external equivalents studies

Three‐phase AC/DC power‐flow for balanced/unbalanced microgrids including wind/solar, droop‐controlled and electronically‐coupled distributed energy resources using radial basis function neural networks

“…Moreover, it should solve this NLES even with different number of equations and operational variables ( n ≠ m ). Here, the solution vector is obtained by solving both the nonlinear PFA (51)–(53) and the operational equation set given by (1)–(50), simultaneously (unlike [8]). The possibility of simultaneous solving of all equations decreases the computation time and provides more accuracy and authenticity.…”

“…The possibility of simultaneous solving of all equations decreases the computation time and provides more accuracy and authenticity. Also, unlike [8], the proposed algorithm can solve NLES even in the event that n > m , and does not need to choose some of the operating variables, arbitrarily. In this paper, the main training parameters are the initial spread and the number of hidden neurons.…”

“…A variety of numerical approaches have been reported for PFA in the technical literature [5][6][7][8][9][10][11][12][13][14]. Nowadays, the intelligent and meta-heuristic algorithms [6][7][8], as well as probabilistic methods [9][10][11], are used to tackle with PFA problems on account of the inherent limitations of classical methods. On the other hand, as the distribution networks and especially microgrids are mostly known to be ill-conditioned power systems due to their weakly-meshed or radial structure and broad-range resistance and reactance values, and higher R/X ratio.…”

“…Newton–Raphson (NR) and its derivatives, such as fast decoupled power‐flow (FDLF), as well as Gauss–Seidel (GS) method, are almost the most familiar approaches for PFA [6]. A variety of numerical approaches have been reported for PFA in the technical literature [5–14]. Nowadays, the intelligent and meta‐heuristic algorithms [6–8], as well as probabilistic methods [9–11], are used to tackle with PFA problems on account of the inherent limitations of classical methods.…”

A New Fast and Efficient Artificial Neural Network Based State Estimator Incorporated into a Linear Optimal Regulator for Power System Control Enhancement

“…To cope with problems 1 and 3, which are more important due to time consumption, the artificial neural network (ANN) approach is introduced to reasonably alleviate them. Neural networks (NNs) have been found in many applications in power system studies, such as load flow studies; distinguishing between transients arising out of capacitor, load, and line switching [15]; short-term and longterm load forecasting; security assessment; state estimation; and many others [16,17]. The feed-forward model in ANNs are capable of approximating any function with a finite number of discontinuities, making it a good candidate in a number of available solutions for solving the power systems engineering problems, specifically estimation processes [9,18].…”