Short-term electrical load forecasting using hybrid model of manta ray foraging optimization and support vector regression

Demand prediction is playing a progressively important role in electricity management, and is fundamental to the corresponding decision-making. Because of the high variability of the increased electrical load, and of the new renewable energy technologies, power systems are facing technical challenges. Thus, short-term forecasting has a crucial utility for generating dispatching commands, managing the spot market, and detecting anomalies. Accordingly, in this work, a hybrid method that optimizes the parameters of a support vector machine using Manta ray foraging optimization is proposed for short-term load forecasting. To evaluate the accuracy of the proposed method, five other optimizers, including the Slime Mould algorithm, Tug of War optimization, Moth Flame optimization, the Satin Bowerbird optimizer, and the Fruit-fly optimization algorithm, are used for evaluation of the proposed method's superiority. By conducting a case study based on actual data, the perfor-mance of all methods is investigated using various statistical indexes. Results show that using the hybridized technique can address the disadvantages and weaknesses of single methods. For example, based on the train and test datasets, the coefficient of determination (R2) values of the proposed SVR-MRFO method for electrical loads are 0.999 and 0.993, respectively.