Joint active and reactive power optimization strategy for active distribution networks based on dynamic community partition

Abstract

With the high penetration of distributed generation, the operational state of active distribution networks exhibits strong uncertainty and fast time-varying characteristics. Power quality issues such as line losses and harmonics have become major factors affecting distribution network optimization. To address this, this paper proposes a joint active and reactive power optimization strategy for active distribution networks based on dynamic community partition. First, causal correlation weights among network nodes are extracted, and the Louvain algorithm is employed for initial community partition. Furthermore, driven by physical constraints, online adaptive updating of community boundaries is achieved. Second, within each community, a multi-objective optimization model that minimizes active power loss and voltage deviation is established, and harmonic constraints are incorporated to actively suppress harmonic pollution while reducing losses and regulating voltage. To solve this problem, a hybrid Tabu Search-Adaptive Particle Swarm Optimization is proposed and embedded into the Alternating Direction Method of Multipliers framework, enabling consistent coordination of boundary variables among communities and distributed parallel solving. Simulation results demonstrate that the proposed strategy effectively reduces system losses and voltage deviations, validating its effectiveness.