comparison between radial basis neural network improvement method with salp optimization algorithm (rbf-ssa) with other hybrid optimization algorithms

In the electricity industry, load forecasting is one of the most important tasks in planning, distribution, operations management, and providing appropriate solutions for power systems. power consumption plays an important role in the planning and optimal use of power systems. with the existing technology, it is not yet possible to store this energy in large dimensions, so accurate forecasting of consumption can play an important role in the economic use of electricity. the amount of electrical charge consumption is not constant but is complex and nonlinearly a function of several parameters. due to the variable amount of electrical charge consumption, power companies must anticipate it in different timelines of the information needed to make decisions. in this article, a new method is presented according to the efficiency of short-term load prediction, which can be from the next few hours to a week or a few weeks. due to the efficiency of evolutionary methods in setting the parameters of forecasting methods, in this paper, the salp optimization algorithm is used as an algorithm with high convergence accuracy to improve the neural network of the radial base function. therefore, in this paper, a comparison between the method of improving the neural network of the radial base function with the salp optimization algorithm for short-term load prediction by considering meteorological factors with other combined methods of optimization algorithms is shown. the results of comparison between predictions in the proposed model (improved neural network with salp algorithm) compared to other combined methods of load prediction, show that the proposed neural network method improves the radial base function with salp (rbf-ssa) better. other combined methods improve the results.