USING THE BEST CHAOTIC ALGORITHM MODEL WITH SUPPORT VECTOR REGRESSION TO PREDICT ELECTRICAL LOAD IN THE REGION TO THE SOUTH

Abstract

The prediction and accuracy of electrical load has received increasing attention, and is one of the important topics in statistics due to the need for it in various areas of life. This process has been the focus of attention of statisticians for a long period of time because prediction has a clear and influential importance in the accuracy of the decision. Most countries in their planning programmers rely on advanced scientific foundations and methods to obtain more effective results. This paper examines a novel and inventive approach for electrical load prediction that enhances prediction performance by combining chaotic hybrid algorithms with super-resolution (SVR) technology. It fixes problems with parameter optimization (SVR). The stability of the support vector regression model (SVR) depends on the selection of ideal parameters. To achieve this, hybridization (SVR) with two chaotic algorithms (CGA, CIA) was conducted. The optimal parameter values in the SVRCIA and SVRCGA models were then compared. The objective was to determine which model was the best and apply it to the estimation of the amount of electrical energy that will be consumed in southern Iraq between 2020 and 2028.

The study concluded that the (CGASVR) model is distinguished and superior to other predictive predictions in terms of statistical prediction criteria.