Efficient Solution of Multi-Period Three Phase Opf Problem Using Spatial Decomposition

The optimized operation of unbalanced power distribution systems with battery-integrated distributed energy resources (DERs) requires the development of efficient multi-period three-phase optimal power flow (OPF) techniques. To overcome the scalability challenges related to the centralized solution strategy, distributed algorithms, especially the alternating direction method of multiplier (ADMM), are largely adopted to tackle three-phase OPF using spatial decomposition. However, ADMM takes many macro-iterations to converge, making it impractical for real-time decision-support. To overcome such deficiencies, in this article, quadratic objective linear programming (QOLP) based Equivalent Network Approximation (ENApp) algorithm is employed to achieve network level optimality by spatially segmenting the centralized OPF problem into smaller subproblems and solving each sub-problem in parallel by enabling boundary variable exchange between the neighbors. Performance evaluations on the IEEE 123 and 2752 bus test cases highlight the effectiveness of the proposed method, achieving substantial reductions in solution times compared to both centralized and ADMM-based methods.