- Literature Review
This comparative analysis confirms the dominance of hybrid AI models (LSTM/GRU ensembles with metaheuristic optimizers) in achieving ultra-low error rates (MAPE 0.06–2%) across forecasting and scheduling tasks. Critical innovations include weather-resilient algorithms, real-time dispatch logic, and wastage reduction architectures
Table1: Summary of Related Works on AI-Based Electricity Load Forecasting and Scheduling