Partial Unobservable Correlation Analysis and Harmonic Source Localization in Active Distribution Networks
Partial unobservability may occur in active distribution networks due to insufficient measuring devices,but with the traditional harmonic source location method it is difficult to ensure the location accuracy of harmonic sources.Based on the sparsity of harmonic sources,a fast sparse Bayesian learning method is proposed to solve the harmonic sources localization for the partial unobservable system.First,by equalizing the unobserved node,the harmonic information of the unobserved node is attached to the unobserved node,and the partial unobserved equivalence model is established and the correlation analysis is carried out.Furthermore,the complex-field sparse Bayesian theory is utilized to solve the equivalence model,and the hyperparameter iteration algorithm based on fast marginal likelihood is employed to speed up the calculation.Finally,the simulation analysis on IEEE33 bus system indicates that the proposed method is superior to the Bayesian compressed sensing algorithm in location accuracy and anti-jamming ability.
partial unobservableactive distribution networkcorrelation analysisfast sparse Bayesianharmonic source location
NETL
NSTL
万方数据
