Transmission Line Fault Location Based On Improved Orthogonal Matching Pursuit Algorithm

Since the reform and opening up,the national economy has developed rapidly.In order to ensure the smoothness of transmission lines and provide sufficient power for all industries in China,the research on the fault location of transmission lines has become increasingly important.Rapid and accurate fault location is the precondition for ensuring the smoothness of transmission lines.At present,due to its high-precision and high-stability advantages,traveling wave method has become one of the objects that electric power workers focus on.In terms of the positioning principle,this paper locates faults by extracting multiple harmonic frequencies of fault traveling waves.First,the functional relationship between the fault traveling wave frequency and the fault location is deduced through the traveling wave trellis diagram.Because the main frequency of the fault traveling wave is greatly affected by the white noise and the line coupling,if only relying on extracting the peak value of a single primary natural frequency for fault location does not guarantee accuracy.In this paper,the fault distance is calculated by extracting the frequency of multiple harmonics.Then the fault location is determined by taking the average of multiple distances.Experiments show that the accuracy of the improved method is improved.In terms of frequency extraction,this paper introduces the Jaccard coefficient to improve the ability of OMP algorithm to improve the recognition ability of similar atoms,and to restore a better signal.Firstly,the advantages and disadvantages of conventional algorithms and the shortcomings of MP algorithm and OMP algorithm in compressive sensing algorithms are analyzed.The resolving power of similar atoms is poor.Then the Jaccard coefficient is introduced instead of the cos coefficient.That is,the arithmetic mean is used instead of the mean value of the set to improve the ability of the algorithm to recognize similar atoms,and a more reliable fault signal can be reconstructed.Finally,the frequency of the reconstructed signal is extracted to locate the fault.Experiments show that the fault signal reconstructed by the improved algorithm introduced with Jaccard coefficients is more realistic,and the frequency of the extracted faults is more accurate in calculating the fault location.In summary,this paper starts from the two aspects of the fault location positioning principle and the frequency of the fault signal,analyzes the existing deficiencies and makes corresponding improvements.From the experimental results,it shows a better reliability and Accuracy to achieve the desired goal.