Abstract
Combination of green energy resources i.e., wind and solar, into electrical power systems is quickly increased in the world. High volatility of such resources made the operation of power system challengeable. For this purpose, short term forecasting is considered as one of the solutions. This solution can be applied to the net demand directly (load forecast minus renewable generation forecast) or it can be applied to the power system indirectly. In this study we have analyzed the proposed two models to show desirable method based on hybrid forecasting approach. The proposed forecasting model consists of three block cascade neural network based on an intelligent algorithm. All parameters of neural network based forecast engine is optimized by intelligent algorithm to increase the prediction accuracy. The proposed forecasting model is then applied on different test cases in various markets and generated results are compared with the results of various prediction models. These comparisons proof the validity of the improved prediction method.