Distributed Collaborative Optimization Model for Multiple Park-level Integrated Energy Systems based on Improved Shapley Value.

Park-level integrated energy system (PIES) is a promising way for efficient utilization of various energy, and the energy exchange among multiple PIESs can better contribute to the energy optimization allocation. However, it is challenging to protect the information privacy and coordinate the interest of different PIESs in the energy sharing process. To this end, a distributed collaborative scheduling model for PIESs based on the improved Shapley value is proposed in this paper. Firstly, integrated demand response is considered in each PIES to fully exploit the flexibility potential of demand-side resources in optimizing energy utilization and satisfaction index is utilized to maintain the user comfort within an acceptable range. Secondly, cooperative game theory is adopted allowing PIESs to exchange energy with others, which can further improve the flexibility and economy of system by taking full advantage of the complementary among different PIESs. Meanwhile, the alternating direction multiplier method is introduced to guarantee their privacy safety. Furthermore, compared with traditional Shapley value, an improved Shapley value benefits allocation scheme with considering the uncertainty of renewable energy and the change in user satisfaction of energy use is established to better avoid the benefits conflicts of all profit-oriented PIESs. Finally, three PIESs with different characteristics are constructed to verify the effectiveness of the proposed model in improving the profits of PIESs, promoting the optimal allocation of energy and coordinating the interest of all stakeholders.