Real-time electricity market transaction mechanism based on Shapley value

As renewable energy sources proliferate, the fairness of incentives provided by the Locational Marginal Pricing mechanism is extreme unfair. We propose a novel approach to distribute trading surpluses using the Shapley value, a method renowned for its mathematically grounded fairness due to its adherence to the axioms of efficiency, symmetry, and marginality. Despite its theoretical advantages, the Shapley value's computational complexity grows exponentially with the number of participants, rendering it impractical for large-scale electricity markets with stringent real-time (5-minute) clearing demands. Our KernelSHAP-based method addresses this challenge by significantly reducing the computational load without compromising the accuracy of the allocation. We have validated our approach on two standard systems and confirmed that it effectively estimates Shapley values, making it a viable solution for fair and efficient surplus distribution in the modern electricity market.