Short-term load forecasting for Jordan’s Power System Using Neural Network based Different

In recent times, there is a burst of interest in load forecasting. It is considered as a backbone activity in the electric power industry. Load forecasting influence decision both from an engineering perspective as well as a financial perspective. The key success in load forecasting is a partnership with the industry and how much the suitable load forecasting method has been widely used by the industry. These methods should be a simple solution to the real world problem. Although some level of sophistication is necessary to achieve high accuracy, so the successful methods typically find a good balance between simplicity and accuracy. Hence neural networks (NN) has received wide concern because of its understandable model, simple and direct implementation and satisfactory performance. In this paper, a short-term load forecasting is demonstrated based on neural networks (NN) adaption in Jordan's power system. Updated parameters are op-timized using different techniques; particle swarm optimization (PSO), genetic algorithm (GA) and elephant herding optimization (EHO). In addition, the error is calculated before and after optimization techniques. As the initial experimental result, the analysis shows that the prediction obtained using metaheuristic optimization based NNs is competitively suitable for the load forecasting. Whereas, two-layer NN gives better prediction with the least error. Therefore, it is clear from the above studies that two-layer NN is the best for load forecasting. Finally, this work is implemented using neural network toolbox and optimization Matlab codes in MathWorks for developing a new and foreseeable future forecasting model.