Accurate models for electric power load forecasting (EPLF) are essential to the operation and planning of a utility company. The load forecasting helps an electric utility to make important decisions, including decisions on purchasing and generating electric power, load switching, and infrastructure development. Load forecasts are extremely important for energy suppliers, financial institutions, and other participants in electric energy generation, transmission, distribution, and markets (He and Zheng, 2018; Jung et al., 2018).

Load forecasts can be divided into three categories (Hsu and Chen, 2003; Papadakis et al., 2003; Al-Kandari et al., 2004):

The forecasts for different time horizons are important for different operations within a utility company. The natures of these forecasts are different as well. For example, for a particular region, it is possible to predict the next day load with an accuracy of approximately 1–3%. However, it is impossible to predict the next year’s peak load with a similar accuracy since accurate long-term weather forecasts are not available. For the next year’s peak forecast, it is possible to provide the probability distribution of the load based on historical load data. It is also possible, according to the industry practice and economic growth to predict the next year’s peak load.

Load forecasting has always been important for planning and operational decision conducted by utility companies. However, with the deregulation of the energy industries, load forecasting is even more important.

In the deregulated economy, decisions on capital expenditures based on long-term forecasting are also more important than in a non-deregulated economy when rate increases could be justified by capital expenditure projects.

Load forecasting methodologies developed may be classified into two broad categories: autonomous models and conditional models. Autonomous models attempt to relate future growth of electricity demand on a system based on its past growth, and conditional models attempt to relate it to other variables, mainly economic indicators.

The load in GCC countries follows almost the same pattern, which is the peak load during the summer season. It is almost doubled the peak in the winter season. So the annual load has been divided into two categories, winter season (low season) and summer season (high season). As stated by Sayed M. Salem (Islam et al., 1995), this will minimize the error and gave a better prediction. After categorizing the load, the percentage increase between 2004 and 2005 for both categories has been calculated. Each category of the year 2005 is multiplied by its relevant percentage, and the same procedure done for the years until 2015. Each load calculated by this technique is added to the historical data and feed to the neuro-fuzzy system to predict the next year.

In recent years, the artificial neural networks (ANN) and fuzzy logic (FL) systems have each providing very encouraging results in solving the problems. This has encouraging researchers to combine both ANN and FL in an attempt to create a final system that reduces the limitations of each of these individual techniques (Tamimi and Egbert, 2000). The strength of this technique lies in its ability to reduce appreciable computational time and its comparable accuracy with other modeling techniques (Metaxiotis et al., 2003) and (Padmakumari et al., 1999). The definition of a neuro-fuzzy system is a combination of ANN and fuzzy inference system (FIS) in such a way that neural network learning algorithms are used to determine the parameters of FIS (Al-Kandai et al., 2003; Mielczarski, 1995). An even more important aspect is that the system should always be interpretable in terms of fuzzy if-then rules, because it is based on the fuzzy system reflecting vague knowledge.

Neural Networks and FL are both complementary technologies in the design of intelligent systems. Each method has merits and demerits. Neural networks are essentially low-level computational structures and algorithms that offer good performance in dealing with sensory data. On the other hand, the FL techniques often deal with issues, such as reasoning, on a higher level than neural networks. However, since fuzzy systems do not have much learning capability, it is difficult for a human operator to tune the fuzzy rules and membership functions from the training data set. Also, because the internal layers of neural networks are always opaque to the user, the mapping rules in the network are not visible and are difficult to understand. Furthermore, the convergence of learning is usually very slow and not guaranteed. Thus, a promising approach for getting the benefits of both fuzzy systems and neural networks is to merge them into an integrated system. This collaboration will possess the advantages of both neural networks (e.g., learning and optimization abilities) and fuzzy systems (e.g., human-like IF-THEN rules thinking and ease of incorporating expert knowledge).

Fuzzy reasoning usually performed using if-then rules. The fuzzy rules define the connection between input and output fuzzy (linguistic) variables. The rules consist of two parts:

$\begin{array}{c}\underset{\text{A Consequence part}}{\underbrace{IF\text{ Month is July and Temperature is High}}}\underset{\text{An Antecedent part}}{\underbrace{THEN\text{ Load is High}}}\end{array}$

For more details, it can be referred in Mielczarski (1995). In the above fuzzy rules, month, temperature level and load are called fuzzy variable and July, high, and high as a linguistic variable. AND is a connective operation, OR is a union and NOT is a complement. It aggregates the results with the premise part.

From a given input to output, using fuzzy logic, Fuzzy Inference (FI) is the actual process of mapping. A number of names know the FIS. These names are: