Analysis And Extraction Method Of Short Circuit Fault Characteristic In Smart Distribution Network

With the integration of the distributed generation(DG),the multi-source of the fault current makes the problem more complex.Traditional distribution network protection can no longer meet the needs of distribution network fault protection.The construction of smart distribution network provides an objective basis for solving the above problems.The development and application of PMU technology in distribution network provides high-precision synchronous phasor data for distribution network,which becomes a reality for centralized fault diagnosis based on multipoint information.To meet the new challenges of multi-source fault current and changeable fault characteristics in smart distribution network,this paper studies the grounding fault characteristics of distribution network with DG.The specific work is as follows:(1)Fault characteristics analysis of distribution network based on distributed power fault model.The main types of DG are introduced.The mathematical models of different types of DG are established based on the analysis of fault output characteristics.Based on the model,the influence of DG type on fault characteristics is discussed.Based on the current situation of various types of DG in distribution network,the influence of multi-types of DG on fault characteristics is analyzed.Simulation models are used to simulate the fault scenario,and the correctness of the above analysis is verified.(2)Fault characterization method based on PMU.The fault characterization method of distribution network based on random matrix average spectral radius index is proposed using the distribution system information system data.The fault characteristic of distribution network based on random matrix theory is proposed to indicate the conditional feature and the result feature of data.The covariance matrix of moving window based on local synchronous phasor information is proposed,and the local covariance matrix is defined.It does not need to store historical data and occupies less memory.Taking the measured data of PMU at each time as samples,a fault feature distribution law based on the improved local covariance matrix is proposed.The feasibility and effectiveness of the proposed fault characteristic are verified by an improved IEEE34-bus example.(3)Fault section location of smart distribution network based on fault feature distribution law.The input matrix is formed by data measured by PMU.Firstly,global monitoring is carried out to judge the occurrence of the failure.The fault region is determined by different variations of the average spectral radius of the matrix combining topology partitioning.The fault section is located in the fault region,and the fault type is judged based on the improved local information covariance matrix.The feasibility and effectiveness of the proposed scheme are verified based on a typical distribution network simulation model with multiple types of distributed generations.