Procedia Computer Science · 2025 · 10 citations · 13 references
ManagementIndustrial EngineeringEnergy ManagementEnergy SystemsMarkov Decision ProcessesDemand ResponseDemand-side Energy ManagementDemand-side ManagementEnergy Management SystemsReinforcement LearningRenewable EnergyDeep Reinforcement LearningEnergy-intensive IndustryIndustrial Demand ResponseIndustrial Energy ConsumptionIndustrial Energy Management SystemsPrice-based Demand ResponseSmart Grid Demand ResponseSmart Grid Integration
Integrating Renewable Energy Resources (RESs) into power grids requires effective Demand Response (DR) programs. Despite high DR potential in industrial sectors, adoption lags behind that of residential and commercial sectors due to diverse operations and production continuity requirements. This paper explores a reinforcement learning (RL)-based DR scheme for energy-intensive industries, promoting the consumption of distributed Renewable Energy (RE) generation. Our approach introduces modifications to the existing Markov Decision Process (MDP) framework. It proposes a flexible reward structure that provides flexibility in balancing production requirements and promotes the consumption of RE. This study addresses the gap in industrial DR literature, emphasizing tailored DR solutions for industrial settings. The key highlight of our RL-based DR solution is its ability to facilitate a price-based DR scheme while promoting the integration of RE into the smart grid.
13
Deep Reinforcement Learning: A Brief Survey
Kai Arulkumaran, Marc Peter Deisenroth, Miles Brundage et al. · IEEE Signal Processing Magazine · 2017 · 4.3K citations
A Survey of Actor-Critic Reinforcement Learning: Standard and Natural Policy Gradients
I. Grondman, Lucian Buşoniu, Gabriel A. D. Lopes et al. · IEEE Transactions on Systems Man and Cybernetics Part C (Applications and Reviews) · 2012 · 1K citations · Full text
Computing, Computer Science, Finance +13