Distributed Optimization Control for Active Distribution Networks with High Penetration of Distributed PV Units

摘要

As the penetration level of distributed photovoltaic system keeps increasing in distribution networks (DNs), overvoltage caused by reverse power flow is an urgent issue to be solved. This paper proposes a distributed optimization control method by scheduling of controllable resources within each partition of DNs. In this optimization problem, the complex relationship between the power flow variables were replaced by a linear formula, and then the power flow model was transformed into a cone programming model. Afterwards, by means of partitioning a DNs into several clusters and decoupling of tie-line in adjacent clusters, the original problem is decomposed into sub-problem of several clusters. According to collected local clusters data and finite boundary data exchange among adjacent clusters, the accelerated synchronous alternating direction method of multipliers (ADMM) are used to iteratively solve the optimal solution of the local clusters. After designing the synchronous ADMM, the convergence rate of synchronous ADMM is improved by the accelerated gradient method. The feasibility and effectiveness of the proposed methods are verified by the case study of the practical DNs in China.