Electricity load and Cost are nitpicking to smart grid action. For different energy direction intention, prediction with various forecasting period of time horizons is reasoned. In our proposed system propose a new method acting regarding the foretelling of generation of renewable energy and electricity load practicing better Bidirectional LSTM. First, a stacked GRU-RNN based training algorithm.It predicts the electricity load and RE by using the training model. Finally, our proposed method verified by foretelling of electrical energy loading with existent energy ingestion data. Dual sensitive observation parameters are chosen reported to the correlativity investigation to shape the input signal data. Secondly, a stacked BILSTM in practice training algorithm, finally it predicts the electricity load by using the training model.Our proposed method verified by foretelling of electrical energy loading with existent energy ingestion data and price. Data-based solution shows that the proposed method using BILSTM achieving an high-fidelity energy foretelling for hard-hitting smart grid operation. © 2024 Elsevier B.V., All rights reserved.