| After the single-phase ground fault occurs in the distribution network,the fault current is weak due to the compensation of the arc suppression coil.At the same time,the distribution network has many branches,complex topology and frequent changes,which challenges the traditional locating methods.With the rapid development of cities,researching a new method that can accurately and reliably locate faults is an urgent problem to be solved.The magnetic field below the overhead line can reflect the changes in line current,and the magnetoresistive sensor has the advantages of non-contact measurement,low cost,and uninterrupted installation and maintenance,which can locate the fault between adjacent towers.By monitoring the magnetic field differences at different positions on the transmission line,this paper proposes new noncontact location methods for grounding faults in distribution networks based on magnetic fields.Based on Biot Savart’s law,the spatial magnetic field beneath the overhead lines was analyzed.Based on the characteristic difference of the upstream and downstream current of the fault point after the single-phase grounding of the ineffective earthing system,the difference of the magnetic field of the x-axis and y-axis in different sections is analyzed.Due to the influence of changes in line structure during magnetic field measurement,this paper explores the differences in magnetic field distribution caused by different tower structures,and analyzes the magnetic field errors caused by line changes and their impact on location methods from two aspects:sag and galloping.In response to the current difficulty in locating distribution networks,this paper constructs three section location methods from different perspectives based on different principles.Firstly,from the perspective of trajectory structure feature analysis,the dual-axis magnetic field of each monitoring point is fitted using the least squares method to obtain the corresponding sine function.The lissajous graph synthesized by the sine function is distinguished from three aspects:major axis,minor axis,and inclination angle,and sub criteria are constructed to achieve section location.Secondly,from the perspective of waveform similarity analysis,five sub criteria,namely relative entropy,F-norm,maximum average difference,cosine similarity,and correlation coefficient,are used to measure the difference in magnetic fields upstream and downstream of the fault point.Based on the calculated results,basic fault probabilities are assigned to each section using D-S evidence theory and further fused,and the section with the highest probability after fusion is selected as the fault section.Finally,from the perspective of microscopic feature quantities,modern signal processing methods are used to mine the fault information contained in the magnetic field waveform.To reduce computational complexity,the mutually perpendicular biaxial magnetic fields are first synthesized,and different frequency band components of the synthesized magnetic field are obtained through variational mode decomposition.Furthermore,the instantaneous energy of the highest frequency mode and the instantaneous phase and polarity of the lowest frequency mode are obtained using the Hilbert Huang transform.The differences in the three instantaneous features are used as the basis for section location.After proposing different location methods,this article analyzes the advantages and disadvantages of each method as well as the applicable scenarios.At the same time,a distribution network simulation model is built in PSCAD to analyze the adaptability of location methods under various operating conditions such as noise,high impedance,different fault initial phase angles,and three-phase imbalance.Furthermore,the impact of sampling data windows and sensor installation positions on the proposed methods is considered.Numerous simulations and actual measurements have shown that the proposed methods has good robustness and reliability under different types of fault conditions. |
PDF Full Text Request