Heap Based Optimization with Deep Learning Based Energy Forecasting in Smart Grid with Consideration of Demand Response

In smart grid, energy management will be an essential one to reduce energy costs of users while maximizing the comfort of users and lessening the peak to-average ratio and carbon emission in realtime pricing techniques. Conversely, the advent of bidirectional transmission and power transfer technologies allows Electric Vehicle (EV) charging or discharging scheduling, load scheduling or shifting, and optimum energy distribution, making the smart power grids. SGs will enable the users to schedule home appliances concerning the Demand Response program (DR) provided by Distribution System Operator (DSO). In this regard, not only the consumers save the cost of using energy, and also it will be very comfortable, but the utility companies also control peakhour demand and diminish Carbon Emissions (CEs). This article introduces a new Heap Based Optimization with Deep Learning Based Energy Forecasting in Smart Grid (HBODL-EFSG) with the consideration of DR. The presented HBODL-EFSG technique majorly focuses on the prediction of energy in SGs by the consideration of DR. To accomplish this, the presented HBODL-EFSG technique applies data standardization process to normalize the input data into a uniform format. For energy-level forecasting, the presented HBODL-EFSG technique uses Deep Variational Autoencoder (DVAE) model. At last, the hyperparameter tuning of the DVAE method was optimally adjusted using the HBO technique. A series of simulation analyses take place to highlight the enhanced forecasting performance of the HBODL-EFSG approach. A comprehensive comparison analysis portrays the precipitated results of the HBODL-EFSG procedure over other techniques.