Privacy-Preserving Outsourced Computation of Collaborative Operational Decisions Among Microgrids in an Active Distribution Network
Xutao Han, Zhiyi Li, Xuanyi Xiao, Ping Ju, Mohammad Shahidehpour
Abstract
This article presents an operational decision-making scheme for facilitating the collaborative decisions between the utility distribution grid (UDG) and microgrids (MGs) in an active distribution network (ADN). The collaborative decision-making among UDG and MGs can help maximize the social welfare of ADN operations, but the decision-making process is faced with ADN computation timeliness and MG privacy leakage concerns. Hence, we propose a proactive deception approach for privacy-preserving outsourced computation of MGs that offers collaborative decisions of UDG and MGs in a centralized and performance-guaranteed manner. First, MG dispatch models are proactively transformed by constructing virtual MGs, which are outsourced to the distribution system operator (DSO), without changing their respective structures and sparsity. Then, the DSO integrates the encrypted MG dispatch models, including virtual objectives, variables, and parameters, with the UDG dispatch model to offer collaborative ADN decisions. Further, the designed asynchronous decryption mechanism will return the DSO's collaborative solutions about points of common coupling which are decrypted in parallel by MG master controllers for updating the distorted UDG power flow. Mathematically, the convexity, feasibility, and correctness of collaborative ADN decision-making are maintained by the proposed approach. Finally, through numerical experiments, the modified IEEE 33- and 123-bus systems are applied to demonstrate the rapidity, accuracy, robustness, and scalability of the proposed approach.