Research on distribution network grounding protect

Research on distribution grounding protection based on modulus maximum theory

sdudy for grouping fault protection in distribution system based on method of the module maximum value

xiao Bai 1,2, Shu Hong Chun 3, Mu Gang 2, LI Ping2

(1．North China Electric Power University , Baoding 071003, China；

2．Northeast China Institute of Electric Power Engineering , Jilin 132012, China；

3．Kunming University of Science and Technology , Kunming 650051, China)

ABSTRACT: Fault circuit selection in distribution system is a problem. In this paper, method of the module maximum value is employed to analyze the electric information after the fault, and to respond the abrupt change of zero sequence currents, and through comparing the change of the zero sequence currents on all lines' polarities and magnitudes, then the fault line can be determined. With the using of scalar product , the criteria has good property of bearing error., and it can get more high precision of fault circuit selection. This method is suited for various distribution system with different neutral grounding modes. Moreover, it can be easily used in site without adjusting protective setting. EMTP simulations show that the method is effective and reliable.

KEY WORDS: distribution system; fault circuit selection; wavelets transform; Single detection

Abstract: in view of the difficulty of single-phase grounding fault line selection in distribution, the wavelet transform modulus maximum theory is applied to analyze the electrical quantity after the fault, reflect the sudden change of zero sequence current, and realize fault line selection according to its different polarity and size change law on each line. The criterion adopts the method of inner product, which has good fault tolerance when processing signals with errors, so it can obtain higher line selection accuracy. This method is applicable to all kinds of neutral point operation modes of small grounding current system, and it is simple to install on site without setting value. EMTP simulation results show that this method is effective and reliable

key words: power distribution; Fault line selection; Wavelet transform; Singularity detection

1 introduction

the neutral ineffective grounding system in which the single-phase grounding arc can extinguish itself is called the small grounding current system [1], which mainly appears in the form of neutral ungrounded, grounded through high resistance and grounded through arc suppression coil. The 3 ~ 60kV power distribution in China usually belongs to small grounding current system

when a single-phase grounding fault occurs in a small grounding current system, a low impedance loop is not formed between the power supply and the fault point, and the short-circuit current is very small. At the same time, the line voltage remains symmetrical, which does not affect the continuous power supply to users, so it is not necessary to trip immediately, and the regulations stipulate that it can continue to operate for 1 ~ 2 hours. However, in order to prevent the further expansion of the fault, it is necessary to select the fault line timely and accurately and remove it

in order to solve this problem, scholars at home and abroad have conducted in-depth and extensive research, proposed a variety of line selection methods based on steady-state components, transient components and external effects (such as: amplitude comparison, phase comparison, harmonic method, compensation method, zero sequence admittance method, power method; first half wave method, energy method, spectral power method, wavelet method; pull method, residual current increment method, injection signal method) [2], and developed corresponding protection devices, Several generations of products have been launched. However, so far, the effect of such devices in actual operation is still unsatisfactory

this paper proposes to apply the wavelet transform modulus maxima theory to find out the change characteristics of electrical quantity after fault, and take the corresponding modulus maxima as the characteristic quantity to analyze, and establish a simple and reliable line selection criterion. A large number of EMTP simulation data show that this method is correct and reliable

2 basic principle

by detecting the singularity of the zero sequence voltage and current in the single-phase grounding fault of the small grounding current system, the polarity and size of their sudden change parts after the fault can be determined, and the fault line can be identified by comparing their different changes on each outgoing line

we call infinitely derivable functions smooth or nonsingular. If a function has a discontinuity somewhere or a derivative of some order is discontinuous, it is said to have a singularity here. Singularity detection is to identify the singularity of the signal and judge its singularity degree. Mathematically, Lipschitz exponent is usually used to describe the singularity of signals [3]. There is a corresponding relationship between the performance of wavelet transform maxima on multiple scales and Lipschitz index [4], which provides a basis for detecting signal singularity and distinguishing singularity through wavelet transform. That is, the modulus maxima after wavelet transform can reflect some characteristics of ground fault, so this method uses this theory to realize fault line selection

2.1 selection of wavelet function

there are infinite kinds of wavelet functions in theory, and the wavelet bases derived from them have different properties, which can meet the needs of various problems. However, different wavelet bases are used to process the same signal, and the results are different, even different. Therefore, in order to obtain satisfactory results, it is necessary to properly select the wavelet function. Although there is no mature method to select the best wavelet function required to solve specific problems, the usual practice is to classify various wavelet functions, summarize the properties and characteristics of each type of wavelet function, determine which type to use in combination with the problem to be solved, and conduct experimental comparison in this type to determine which wavelet function to use [5]

as mentioned above, for the specific problem of fault line selection in small grounding current system: in order to reduce the leakage and aliasing of the spectrum, the wavelet function is required to have good frequency domain characteristics. DBN wavelet system is a wavelet function widely used in engineering. The characteristics of this wavelet system are that with the increase of sequence number n, the time domain support becomes longer and the time locality becomes worse; At the same time, the regularity increases and the frequency domain locality becomes better. However, when n increases to 10, the frequency division performance of DBN wavelet in the frequency domain is very close to that when n is 10

considering the needs of analysis in the time-frequency domain and the characteristics of fault line selection, several wavelets are used for multiple simulation calculations, which proves that using db10 wavelet can get ideal results. Therefore, db10 wavelet is selected in this paper, and the waveforms of its scaling function and wavelet function are shown in Figure 1 (a) and (b) respectively

2.2 criteria for line selection

first, wavelet transform is performed on the data of zero sequence current on each outgoing line within one cycle before the fault and three cycles after the fault to obtain a corresponding set of modulus maxima, where n represents the line number, and I represents the sequence number of the current maximum value with the financial support of the Federal Ministry of education and research of Germany. Then, arbitrarily select an outgoing line as the reference line, and make the inner product of the wavelet transform modulus maxima group of zero sequence current on it and the wavelet transform modulus maxima group of zero sequence current on other lines, and take this inner product result as a measure, expressed in S

in formula (1), j is the number of the arbitrarily selected reference line; K is the number of the remaining lines, that is, k = 1, 2,... N, and K ≠ J; N is the total number of outgoing lines; M is the number of modular maxima

in this way, the following line selection criteria can be established:

(1) if SJK is not greater than zero or less than zero at the same time, the established line is a non fault line; The line that makes it established is the fault line

(2) if SJK is less than zero at the same time, line j is a faulty line

(3) if SJK is greater than zero at the same time, it is a bus fault

2.3 description of line selection criteria

first, due to the algorithm of wavelet transform itself, the right boundary of the data window will be regarded as a sudden change point in the transformation process, so that each scale component will have a large value near the right boundary, which is the boundary effect of wavelet transform. In order to overcome the adverse effect of boundary effect on route selection, only the maximum value in the first two cycles is taken as the inner product

secondly, the essence of inner product is polarity comparison. The modulus maxima with large amplitude is beneficial in the comparison process and the result is reliable; The modulus maxima with small amplitude will be easily affected by errors in the comparison process, so as to get wrong conclusions. By doing inner product, it is equivalent to making the comparison result of the larger amplitude occupy a high weight in the measure, while the comparison result of the smaller amplitude occupy a low weight in the measure. In this way, the influence of error is overcome to a great extent, and the accuracy of line selection is improved

again, wavelet singularity detection reflects the singularity of the signal and does not require the signal to be jump [6]. Therefore, although this method uses the numerical value in the transient process to analyze, this method is still effective when a single-phase grounding occurs near the zero crossing of phase voltage

in addition, since this method is a line selection method based on transient components, although the grounding line can be selected instantaneously in actual use, in order to distinguish between transient faults and permanent faults, it is also necessary to judge whether the fault still exists after a time delay before deciding whether to perform tripping operation

3 simulation analysis

simulation analysis is carried out for a 35KV radial small grounding current system (as shown in Figure 2) when the neutral point operation mode is grounded through arc suppression coil. Incidentally, this method is also applicable to systems with ungrounded neutral and high resistance grounding

assuming that phase A is grounded at 0.315 seconds 24 kilometers from the beginning of line 4, taking the transition resistance of 1 ohm and the sampling rate of 10 kHz as an example, the line selection is realized according to the above method. Due to space limitation, only the analysis waveforms of line 2 (normal line) and line 4 (fault line) are given, as shown in figures 3, 4 and 5

here, line 1 is selected as the reference line, and the modulus maxima measures of zero sequence current on lines 2, 3, 4 and 5 are 351.1, 540.7, -1200.5 and 216.8 respectively. According to the above criteria, line 4 is a fault line

in order to facilitate comparison, under the condition that the transition resistance, sampling rate and reference line are the same as before, this method is used to simulate and analyze the system shown in Figure 2 when the short line near end, the short line far end, the long line near end, the long line far end and the bus is grounded. The obtained wavelet transform modulus maximum measures are listed in Table 1. As line 1 is a reference line, its measure is the result of inner product with its own wavelet transform modulus maxima (must be positive), so the wavelet transform modulus maxima measure of the line does not need to be calculated, which is represented by "+" in the table. In this way, the fault line can be selected very accurately by analyzing these data according to the above line selection criteria

take the system shown in Figure 2 as an example. When the sampling rate is still 10 kHz, the transition resistance increases to 2000 ohms, and the reference line becomes outlet 2, the method is further tested. The simulation data are shown in Table 2. The numerical value, on the one hand, shows that the reference line can be selected arbitrarily, and the correct line selection results can also be obtained; On the other hand, it shows that this method has strong resistance to transition resistance

4 conclusion

because this method uses the data of several cycles near the fault point to realize line selection, at this time, the change of electrical quantity is usually obvious, and the amplitude of characteristic quantity is large, so it has high line selection accuracy. At the same time, wavelet singularity detection reflects the singularity of the signal