Contents lists available at ScienceDirect Electrical Power and Energy Systems journal homepage: www.elsevier.com/locate/ijepes Optimum control strategies for short term load forecasting in smart grids Mansoor Ali, Muhammad Adnan, Muhammad Tariq ⁎ National University of Computer and Emerging Sciences, Pakistan ARTICLEINFO Keywords: Short term load forecasting Power fow control Overloading Fuzzy control ABSTRACT Nonlinearity in load profle and variations in demand due to error margin in short term load forecasting cause power network overloading. The state of a power system is more severe when a fault occurs in the power system network that leads to overloading. Analyzing the efect due to these disturbances on power system network is an important feature of this work. This paper proposes a control algorithm that focuses on sophisticated fuzzy logic approach. Advanced fuzzy control takes overloading and variation in demand profle as input, which mitigate these disturbances by in- corporating optimal power dispatch of renewable energy resources (RERs). To show the efectiveness and validity of the proposed model and fuzzy control design, 9 Bus test system of the transmission network is adopted. Not only normal mode but fault and overloading modes are used to verify the proposed approach. Competitiveness of the proposed control design in terms of reliability and optimal utilization of RERs are verifed through simulation results. 1. Introduction Electric power infrastructure is the backbone for every country and is an important factor that directly afects the economic policy of a country. The traditional conventional electric power grid is not advancing in terms of control and reliability. The era is now moving toward the smart grid that incorporates advanced sensing, communication, security, and control technologies, which make a grid more reliable and efcient [1]. One of the most important features of the smart grid is that it gets power from diferent type of Distributed Generation (DGs) sources in order to meet the power demand at a cheaper cost. Besides providing cheaper power, there are some drawbacks associated with DGs and an important one is reliability [2]. If a power grid is totally supported by renewable energy resources (RERs), it leads to serious overloading failure because of limited load handling capability of RERs. In order to meet demand profle and low power losses in the power system, electrical load forecasting is a very important factor for utilities and power system operators. Many operating decisions such as eco- nomic dispatch of the power plant, designing of the power network and security network depend upon load forecasting. Electrical load fore- casting mainly consists of four types: very short term, short term, medium term, and long term. The short term load forecasting (STLF) is mostly done for duration varies from hours to weeks. New advanced technologies are introduced for monitoring of demand response profle and integration of generation sources in smart grids. These technologies use intelligent and adaptive elements that require more advanced techniques to perform accurate generation and demand forecasting in order to work optimally. The authors in [3] briefy defned various types of load forecasting, which can be efciently utilized in a power system network. Similar to very short term load forecasting (VSTLF), which is used for power fow control, STLF is used for the adjustment of the generation and demand, where as medium term load forecasting (MTLF) and long term load forecasting (LTLF) are used to plan assets’ utilities. Load forecasting is further classifed into two groups in [3] frst group is used to forecast single value while second group is used to forecast multiple variables. Authors in [3] also mentioned diferent forecasting techniques their respective accuracy and in which scenario they will be more useful. Authors in [4] forecasted aggregated load using artifcial neural network (ANN) while taking in account diferent variables that afects the forecasted aggregated load. Among diferent variables that are analyzed the most important one is the climate. Diferent testing model are suggested in [4] and some of the model include climate as input variable in order to show how forecasted ag- gregated load is afected by climate variable. STLF is an important factor used in determining of power plants’ work plan and choosing of best production group. Energy companies face many economic and technical problems such as operation, planning, and control of power system network. Decisions about energy production, infrastructure de- velopment, and load switching are made easier for utility companies through STLF. Therefore, forecasting a load correctly is an important factor in the competitive market for utility companies [5,6]. The elec- trical utilities must manage the supply from generation sources in order to meet the demand of its consumers. It is therefore very important for the utilities for having advance knowledge of their related consumers https://doi.org/10.1016/j.ijepes.2019.06.010 Received 22 December 2018; Received in revised form 30 May 2019; Accepted 6 June 2019 ⁎ Corresponding author. E-mail address: tariq.khan@nu.edu.pk (M. Tariq). Electrical Power and Energy Systems 113 (2019) 792–806 0142-0615/ © 2019 Elsevier Ltd. All rights reserved. T