ORIGINAL ARTICLE Flower pollination–feedforward neural network for load flow forecasting in smart distribution grid Gaddafi Sani Shehu 1 • Nurettin Çetinkaya 2 Received: 11 October 2017 / Accepted: 3 March 2018 Ó The Natural Computing Applications Forum 2018 Abstract Nature-inspired population-based metaheuristic flower pollination algorithm is proposed in solving load flow forecasting problem in smart distribution grid environment. The efficient approach involves training a feedforward neural network (FNN) with a new flower pollination algorithm (FPA). The idea is to perform short-term load flow forecasting in smart distribution network, thus maintaining system security due to intermittency of renewable energy penetration and power flow demand. Application of optimization algorithms such as FPA in training neural network improves accuracy, over- comes generalization ability of neural network, requires less data and prevents premature convergence problem in artificial intelligence solutions due to nonlinearity of parameters. The real load flow data are collected through distribution man- agement system of Konya Organized Industrial Zone. The result obtained indicates strong improvement in error reduction using flower pollination optimization algorithm in training FNN for short-term load flow forecasting in smart distribution grid; the model is compared against FNN model and efficient support vector regression. Keywords Flower pollination algorithm  Feedforward neural network  Load flow forecasting  Smart distribution grid 1 Introduction Over the past few decades, electricity sector is confronted with growing energy demand and high level of renewable energy penetration and couples with total deregulations of power sector as commodity enterprises. In other hand, high aggregated power losses and power flow forecasting con- stitute critical challenges to the industry. These challenges lead to development of new power system technology called smart grid; the new grid is to ensure efficiency and sustainability in meeting electricity generation, transmis- sion and distribution demand with reliability and best of quality at optimal cost. The term smart grid has been coined to define the present day development of power system, accommodation of renewable energy integration, user’s activities to the network, and those that do both in order to deliver efficiently sustainable, economic and secure electricity supplies [1]. As such other innovations focusing on connecting low-voltage meters to be remotely controlled are raising new chances for load flow forecast- ing. Currently short-term load forecasting is a major col- umn in everyday life of power networks for maintaining energy demand and component system security. An accu- rate prediction is necessary to issue short- and long-term network operation plans. Hence, any inaccuracy or devia- tion may result in a loss of significant power in MW or million amount of operational cost for the utility company [2]. Load flow is performing to ascertain power system economic operations and component security. For optimal utilization of power system, state of operation ought to be well known in the present and several hours or day onward to overcome security margins and plan equipment maintenance. Forecasting method is popular among many researchers especially in power system load demand, several & Gaddafi Sani Shehu gsshehu@selcuk.edu.tr Nurettin C ¸ etinkaya ncetinkaya@selcuk.edu.tr 1 Graduate School of Natural and Applied Science, Selc ¸uk University, Alaeddin Keykubat Yerles ¸kesi, Akademi Mah. Yeni I ˙ stanbul Cad. No: 369, 42130 Selc ¸uklu, Konya, Turkey 2 Electrical and Electronics Engineering Department, Selc ¸uk University, Alaeddin Keykubat Yerles ¸kesi, Akademi Mah. Yeni I ˙ stanbul Cad. No: 369, 42130 Selc ¸uklu, Konya, Turkey 123 Neural Computing and Applications https://doi.org/10.1007/s00521-018-3421-5