Application of weighted least square algorithm in distribution network state estimation with finite measurement information

摘要

In order to adapt to the requirement that distribution network has limited measurement devices; a distribution network state estimation algorithm based on finite measurement information is proposed. In the model, distribution network node voltage amplitude and voltage phase work as state variables, and measured node information and load planning data work as real measurement and pseudo measurement respectively. The weight matrix helps ensure the model's rationality and solving model is based on the weighted least square algorithm. With the simulation of UK 102-node distribution network, the results show that this model could estimate all-node states with finite measurement information and mean error meets the requirements of actual application. The relation of state estimation error, real-measurement number and estimation initial value has qualitative description, and it lays a foundation of further study on optimized measurement configuration and optimal state estimation solving.