Fault Line Selection And Fault Section Locating Based On S-Transform For Single-phase-to-ground Fault In Distribution Networks

In China, the neutral ineffectively grounded system (called low-current grounding power system as well) is widely used in the medium and low voltage distribution networks. This kind of grounding mode has the advantages of improving power supply reliability. But, when single-phase-to-ground fault happens, it is difficult to locate accurately because of the weak current in fault point. At present, the study of fault location is mostly focused on the fault line selection at home and abroad. But it is lack of fairly mature research on how to fix the fault segment. Therefore, it is still a problem to be solved urgently on how to locate rapidly and accurately after fault.Comparing with steady-state signal after fault, the transient signal has the characteristics of much greater amplitude than steady-state signal and it doesn’t be affected by arc suppression coil and unstable arc in fault point. The transient signal can effectively make up the shortcomings of fault location by using steady-state signal. The existing transient location method only analyzes the time domain characteristics of transient signal, ignoring the difference between different frequencies. The frequency domain characteristics of transient signal as well as extracting and using of fault information are lack of researching.Aiming at these problems above, the paper proposes a fault location method that using S-transform to extract the dominant resonant frequency of transient zero module current after fault, and then realizing the single-phase-to-ground fault location quickly and accurately with intelligent distribution terminal. The main research works are as follows:(1) The frequency domain characteristics of transient zero module current are analyzed by an accurate faulty model based on distributed parameters in this paper, mainly about the distribution law of transient zero module current in a specially selected frequency band (SFB). Usually, the dominant resonant frequency of transient zero module current in upstream of fault point is within SFB limit, while it is out of SFB limit in non-fault lines and fault point to load. Within the SFB, the equivalent impedances of all feeders in zero module networks are capacitive, and each can be modeled as a lumped capacitor. The signals within SFB have well defined characteristics and contain most energy of transient zero module signals which can ensure the reliability and accuracy of detection. The signals within SFB can be used as the basis for fault location.(2) A method of using S-transform to extract the dominant resonant frequency of transient zero module current is presented in the paper. S-transform has good time-frequency resolution and time-frequency location capability. It can reflect partial feature of non-stationary signal. S-transform is both suitable for analyzing non-stationary signal with mutation, and has advantages of visual result and explicit physical meaning in feature extraction. Therefore, S-transform is suitable for extracting transient zero module signals.(3) The basis on single-phase-to-ground fault line selection and fault location is put forward by analyzing frequency domain characteristics of transient zero module current combined with S-transform. When fault happens, the short window data of first1/4cycle are selected. The S-transform is carried out to determine the dominant resonant frequency of transient zero module current. Then the fault location is determined by comparing the dominant resonant frequency of transient zero module current in each detection point with SFB. Combined with intelligent distribution terminal, this paper also put forward a fault location method based on intelligent distribution terminal.(4) Using Electromagnetic Transients Program EMTP-ATP, the simulation models of small current grounding system are built and a lot of experiments of single-phase-to-ground fault are carried out. Through the simulation data extraction, the fault location method based on S-transform is verified by using MATLAB. The simulation results proved the correctness and effectiveness of this fault location method.