Microgrid secure pricing and real-time flexible energy market integrated with trust-based demand response: a multi-layer FDRL architecture

• Secure, resilient, and adaptive energy market control in microgrids. • Cyber-resilient control strategy for real-time pricing in flexible energy markets. • Trust-aware demand response functionality to manage participation incentives. This article presents a novel cyber-resilient control strategy for enabling secure, real-time pricing in flexible energy markets of microgrids. As the energy infrastructure gets more digitally connected and more integrated with distributed energy resources, microgrids face increased challenges to their capacity to preserve pricing integrity, system trust, and operational resilience in the face of evolving cyber threats. To address these issues, this paper developed a federated deep reinforcement learning model based on the soft actor-critic algorithm. The system includes long short-term memory-based anomaly detection for early cyberattack detection, trust-aware demand response (DR) functionality to manage participation incentives, and a privacy-preserving coordination scheme to enable decentralized learning without raw data exchange. This multi-layered configuration supports local agents in making optimal joint pricing and control decisions and being responsive to false data injection, denial of service, and hybrid cyberattack scenarios. Simulations in IEEE 33-bus, 69-bus, and 141-bus power systems corroborate that the introduced model maximizes social welfare by up to 11%, reduces coordination failures by over 50%, and maintains DR compliance over 92%. The price volatility index is below 0.02 under attack. These results demonstrate the framework’s capability of enhancing economic stability, cyberattack resilience, and scalable coordination in secure energy market settings.