Optimization and Leveraging of Peak Demand Using LSTM and Fuzzy Logic Approaches in Demand-Side Management

This paper explores the application of Fuzzy Logic and Long Short-Term Memory (LSTM) to Demand-Side Management (DSM) peak load demand forecasting, addressing the critical need for more accurate predictions in optimizing energy use and reducing costs. While LSTM uses historical load data to capture complicated temporal correlations, Fuzzy Logic efficiently handles uncertainty. Further Random Forest technique applied to the same data which fails to give impressive result. By combining these approaches LSTM outperforms Fuzzy Logic and Random Forest, notably reducing peak demand. It improves interpretability and computational efficiency for real-time DSM applications. With an emphasis on transparency in DSM decision-making, the primary contribution is novel hybrid model that increases predicting accuracy. In order to improve DSM capabilities even more, future research areas support the use of hybrid models and sophisticated AI.