Sparrow Search Optimization with Deep Belief Network based Wind Power Prediction Model

Wind power is a clear feature of intermittent, nonstationary, and difficult fluctuations, making it challenging for achieving consistent wind power generation. Assuming the restricted nature of typical energy resources and the developing difficulties of environmental problems, several countries are starting with developing novel energy resources which are considered for environmental and renewable safety. Amongst the several novel energy resources, wind energy was abundant, doesn’t cause pollution, has minimum cost, and does not deplete. Accurate wind power predictive is enhance the reliability and safety of power grid function. Therefore, this study presents a sparrow search optimization with deep belief network for wind power prediction (SSODBN-WPP) technique. The SSODBN-WPP technique follows a two stage process namely prediction and parameter tuning. At the initial stage, the SSODBN-WPP technique employs DBN method for wind power prediction. Next, the SSO algorithm is used to adjust the core hyperparameters of the DBN algorithm. The efficacy of the SSODBN-WPP method is tested on a comprehensive set of simulations that take place on wind power dataset. A comparison study of the SSODBN-WPP technique reported its betterment over other predictive approaches.