Research On Fault Diagnosis Of Double-ended Cable Based On The Wavelet Singular Entropy And Algorithm Fusion

With the expansion of the scale of the city and the continuous extension of the power lines,the overhead lines in urban power grid are gradually replaced by cable lines.Once the Underground cable fault happens,it will cause a power outage,and bring economic losses to the factory production,but also bring a lot of inconvenience to people's life.Quickly troubleshooting and restoreing the power supply are important tasks to ensure the stable operation of the power system,and the key to the troubleshooting is the fault recognition and positioning.For this purpose,this paper had conducted the thorough research about the method based on wavelet entropy and algorithm fusion of double-ended cable fault diagnosis,the main contribution had the four aspects:(1)Taking the wavelet singular entropy as the starting point,and applying the wavelet singular entropy to the fault phase selection of the cable line in the distribution network,and a line selection algorithm based on the transient quantity of fault voltage was proposed.The results of simulation and system experiment showed that the proposed algorithm had a good recognition effect,and the false positive rate is only 0.74%,and it was not easily affected by different transition resistances,faults point and faults occurring time.It had a high sensitivity and adaptability.(2)In order to eliminate the inter-line coupling between the three phases of the cable lines,the originally collected fault signal at the detection point is decoupled,and an improved phase-mode transformation matrix was proposed.It can realize the characterizing of all fault types with a single modulus,and overcome the shortcomings that single modulus of the existing phase-modulus transformation matrix can not reflect all types of faults.(3)In order to eliminate the pseudo-root problem caused by double-ended traveling wave fault location,a fault detection algorithm was proposed based on improved parameter detection.The algorithm used data of the voltage and current at both ends of the fault to correct the asynchronous angle and the line parameters in real time,and then used the fault data to finish the preliminary fault location.The proposed algorithm did not have the problem of pseudo root discrimination.(4)In order to eliminate the influence of the uncertainty of the wave velocity on the accuracy of the traveling wave fault,a fusion method of double-ended traveling wave ranging from improved parameter detection and wave-wave was proposed.The complex wavelet transform was used to analyze the traveling wave signals,and got the three wavehead times of the initial wave of the fault,the reflection point of the fault point,and the reflected wave of the opposite side of the reflection of the fault point,and derived the fault ranging formulas independent of the wave velocity.The results of simulation and system experiment showed that the ranging error is less than 4%,and the method was not affected by the fault type,the transition resistance and the fault time.Therefore,the research of this paper had good scientific and practicality for power cable fault diagnosis.