Electricity supply chain hybrid long‐term demand forecasting approach based on deep learning—a case study of basic metals industry

Demand forecasting is a key parameter to achieve optimal supply chain management at different levels. The basic metals industry one of the most energy-intensive industries in electricity chain. impossibility large-scale energy storage, reservation constraints, and its high costs, limitations on transmission lines capacity, real-time response priority strategic demands, variety rates hours seasons are issues that challenge A hybrid approach presented this paper improve accuracy long-term demand proposed uses wavelet decomposition long short-term memory (LSTM) neural networks produce new predictors for case defective data. data used study consists recorded hourly Espidan Iron Stone (EIS) company Mobarakeh Steel (MS) Isfahan Province. understudy time series includes lot interruptions due non-production factory or power outages spikes which cause uncertainty makes more difficult comparison with continuous series. Multiple machine learning models including decision trees (DTs), boosted bagged (BGTs), support vector regression (SVR) models, extreme machines (ELMs), LSTM network model employed evaluate effectiveness paper. results obtained using training (ML) methods shows notable reduction error.