基于状态空间线性变换的主动配电网分布式电压控制

Distributed voltage control of active distribution network based on state space linear transformation

大规模分布式发电的接入使得主动配电网电压控制问题日益凸显。受限于配电网模型参数的精度,传统集中式的电压控制和基于模型的分布式电压控制策略效果受到显著影响。提出了一种基于状态空间线性升维变换的主动配电网分布式电压控制方法。通过矩阵分裂方法实现了海森矩阵的分布式求逆,将分布式控制收敛速提升至超线性收敛。基于Koopman数据驱动方法,利用配电网历史运行数据作为训练样本,构建高维线性精确潮流模型,从而推导得到电压-无功全局灵敏度,以此校正分布式牛顿控制中的迭代方向。算例结果证明,相比依赖于模型的分布式电压控制方法,所提方法具有更快的收敛速度和更优的控制收敛结果,且不受参数不精确问题影响,具有更强的工程适用性。

The integration of large-scale distributed generation into active distribution network poses challenges to voltage control. Limited by the accuracy of distribution network model parameters,the performance of traditional centralized control and model-based distributed control is significantly affected. Aiming at the above problem,a distributed voltage control of active distribution network based on state space linear transformation is proposed. By utilizing matrix splitting method,it can carry out Hessian inverse in a distributed manner,and provide super-linear convergence. Based on Koopman data-driven method,the historical operation data of distribution networks is taken as training samples,the lift-dimension linear power flow model is constructed,and the voltage-reactive power sensitivity is derived. Therefore,the Newton direction in distributed control can be properly tuned. The results of case studies validate that compared with the methods based on model parameters,the proposed method exhibits faster convergence rate and better voltage profile. Besides,the method is independent on parameters,and has superiority in practical applications.