Research On Distribution System State Estimation In Active Distribution System

With the urgent needs of the energy crisis and environmental issues and rapiddevelopment of distributed generation (Distributed Generation, DG) technology,that drives more and more DG access to the distribution network. DG’s accessmakes the distribution network changing from a single power supply system to amulti-ring weak radial network system, a reverse trend and voltage violationsproblems on the feeder, plus random fluctuations of load and DG outputfluctuations, have posed a serious challenge to the distribution network operation.Distribution System State Estimation(DSSE) as Distribution Management System(DMS) core part of the distribution network can provide high reliability data foradvanced management software, such as voltage regulation control, DG poweroutput control and load allocation for the distribution network security and stableoperation. Due to lack of development of the distribution network innate,Distribution System State Estimation still faces many challenges: the lack ofeffective real-time measurement, a huge asymmetry with detailed modeling gridnetwork structure and the ability to adapt to the three-phase unbalanced state withDG estimation algorithm. Therefore, this article does research on distributionsystem state estimation with distributed generation from the three aspectsmentioned above.Firstly, for the three-phase unbalanced distribution network characteristics,OpenDSS platform is to build a more accurate model of three-phase powerdistribution network, and propose measured and unmeasured state estimationmodel based on different measurement configurations of DG; and then based on thesystem of built three-phase model, we propose flow matching state estimationalgorithm considering VR characteristics of the distribution network with DG; thealgorithm, intergrated the proposed pseudo load measurement the weightoptimization algorithm based on GMM model to solve the unbalance of the powerdistribution coefficient matching process optimization problem. Finally, throughOpenDSS/MATLAB cross-platform programming proposed models and algorithmsand through concrete examples to verify the validity and rationality of the proposedmodels and algorithms.respectively.