Data-driven pricing and trusted trading mechanism for park-level flexible resources under energy-carbon coupling

Abstract

The coupling of electricity and carbon markets presents new challenges for the operation of industrial park integrated energy systems. This paper proposes a comprehensive framework to address inaccurate response modeling, scalability limitations in pricing, and a lack of trust in distributed trading. A dynamic sensitivity matrix based on a hybrid neural network is developed to quantify the real-time response potential of flexible resources. A three-layer Stackelberg game model is then constructed and solved via a distributed algorithm to determine optimal pricing. Furthermore, a blockchain-based verification mechanism is introduced to ensure transaction trust. Validation based on real-world data from an industrial park in the Guangdong-Hong Kong-Macao Greater Bay Area demonstrates that the proposed method effectively captures energy-carbon coupling effects and improves market efficiency and fairness.