A Fault Identification Method Based on Line Forward-Reverse Traveling Wave Correlation

Abstract

The present invention relates to a fault identification method based on the correlation of forward and reverse traveling waves of a line, comprising: S1, collecting real-time voltage and current signals of a multi-terminal flexible DC power grid system; S2, calculating the line mode voltage based on the real-time voltage, obtaining the line mode voltage differential value as a starting criterion, returning to S1 if the line mode voltage differential value is not greater than the preset voltage value, and entering S3 if the line mode voltage differential value is greater than the preset voltage value; S3, fitting the current signal, obtaining the current change trend, and judging the fault within the area and the forward out-of-area fault and the reverse out-of-area fault based on the current change trend; S4, obtaining the similarity of the voltage traveling waves on this side and the opposite side, judging the fault within the area and the forward out-of-area fault based on the similarity, and selecting the fault pole by using the ratio of the voltage mutation amount after the fault when the fault is an internal fault. The present invention does not require a complex calculation process, and only needs to extract the voltage traveling waves at both ends of the fault, and the algorithm has good reliability.

Description

Technical Field

[0001] The invention relates to the technical field of DC line fault identification, and in particular to a fault identification method based on line forward-reverse traveling wave correlation. Background Art

[0002] At present, the protection schemes for flexible DC grids can be roughly divided into protection schemes based on single-terminal information volume and protection schemes based on double-terminal information volume.

[0003] The single-ended quantity protection scheme is a protection scheme constructed based on the electrical quantity information of the single end of the protection line. Its characteristics are that no information interaction is required and the speed is good. Due to the characteristics of the development of DC line faults, the speed requirement is high, so the single-ended quantity protection is often used as the main protection of the DC line. The current single-ended quantity protection schemes can be mainly divided into traveling wave protection, differential protection, protection based on boundary characteristics, distance protection, and protection based on artificial intelligence algorithms. Since there are many cases where single-ended quantity protection has a protection dead zone at the protection end, the fault quantities at both ends are required for fault detection. The protection scheme based on double-ended quantities has higher reliability and sensitivity, and can usually be divided into longitudinal differential protection and directional longitudinal protection.

[0004] With the rapid development of flexible DC technology, the structure of DC grid system is becoming more and more complex, and line fault identification is becoming more and more difficult, which is mainly reflected in the following aspects:

[0005] (1) Compared with traditional AC systems, faults in flexible DC systems develop faster and the fault process is shorter. The fault current rises to its peak value within a few milliseconds. Traditional AC power grid line fault identification methods are difficult to meet the speed requirements in DC systems.

[0006] (2) Compared with conventional high-voltage DC systems, conventional DC grids are generally two-terminal structures, and there is no need to consider other faults outside the line area when designing protection schemes. However, flexible DC grids tend to develop into multi-terminal ones, which puts higher requirements on the speed and selectivity of fault identification. The fault identification method of conventional DC may no longer be applicable in multi-terminal flexible DC grids. Summary of the invention

[0007] The purpose of the present invention is to provide a fault identification method based on the correlation of forward and reverse traveling waves of a line. Based on the voltage traveling wave relationship at both ends of the line, a method for identifying the inside and outside of the zone based on the similarity of the traveling waves at both ends is proposed. First, the mutation direction of the current is fitted according to the least square method to judge the fault inside the zone and the forward outside the zone fault and the reverse outside the zone fault. Then, according to the judged fault inside the zone and the forward outside the zone fault and the reverse outside the zone fault, the corresponding forward traveling wave and reverse traveling wave are taken to calculate the Pearson correlation coefficient to identify the fault inside and outside the zone. The fault pole is selected by the ratio of the voltage mutation amount between the positive pole and the negative pole. The identification method does not require a complicated calculation process and only needs to extract the voltage traveling waves at both ends of the fault. The algorithm has good reliability.

[0008] To achieve the above object, the present invention provides the following solutions:

[0009] A fault identification method based on line forward-reverse traveling wave correlation, comprising:

[0010] S1, collecting real-time voltage and current signals of a multi-terminal flexible DC grid system;

[0011] S2, calculating the line mode voltage based on the real-time voltage, obtaining the line mode voltage differential value as a start criterion, returning to S1 if the line mode voltage differential value is not greater than the preset voltage value, and entering S3 if the line mode voltage differential value is greater than the preset voltage value;

[0012] S3, fitting the current signal to obtain a current change trend, and judging the fault within the zone and the forward fault outside the zone and the reverse fault outside the zone based on the current change trend;

[0013] S4. Obtain the similarity of the voltage traveling waves on the local side and the opposite side, and judge the fault within the zone and the fault outside the positive zone based on the similarity. When the fault is an internal fault, select the fault pole by using the ratio of the voltage mutation amount after the fault.

[0014] Optionally, judging the fault within the zone and the forward out-of-zone fault and the reverse out-of-zone fault based on the current change trend includes:

[0015] Fitting the current signal according to the least square method to obtain the current change trend;

[0016] An identification criterion is constructed according to the slope of the current change trend, and the in-zone fault, the forward out-zone fault, and the reverse out-zone fault are determined according to the identification criterion.

[0017] Optionally, the identification criterion is:

[0018]

[0019] Among them, S is the pole selection criterion, sgn is the sign function, and k is the slope of the current change trend.

[0020] Optionally, obtaining similarity between the voltage traveling waves on the local side and the voltage traveling waves on the opposite side, and judging the fault within the zone and the fault outside the forward zone based on the similarity includes:

[0021] The reverse wave of the protection line voltage at the preset time after the fault is extracted and the forward wave of the protection line voltage at the opposite side is calculated by Pearson correlation coefficient;

[0022] The in-zone fault and the out-of-zone fault are judged according to the Pearson correlation coefficient.

[0023] Optionally, the method for judging the intra-zone fault and the positive out-zone fault according to the Pearson correlation coefficient is:

[0024]

[0025] Among them, ρ is the correlation coefficient of the voltage traveling waves on both sides of the line, ρ set is the threshold value that is set.

[0026] Optionally, when the fault is an intra-zone fault, the fault pole is selected by using the ratio of the voltage mutation amount after the fault, including:

[0027] Obtain voltage data with a time window of a preset time after the fault, and calculate the ratio of the positive and negative voltage mutations;

[0028] The faulty pole is selected based on the ratio of the voltage mutation amounts of the positive pole and the negative pole.

[0029] Optionally, the method for selecting the fault pole is:

[0030]

[0031] Among them, P set represents the setting threshold, and P is the ratio of the positive and negative voltage mutations calculated from the 2ms data after the fault.

[0032] The beneficial effects of the present invention are as follows: (1) The present invention does not require a complicated calculation process and only needs to extract the voltage traveling waves at both ends of the fault. The algorithm has good reliability. During calculation, only 2ms data after the fault needs to be extracted for processing, and there is no need to consider communication synchronization issues. (2) The present invention has strong resistance to transition resistance and noise interference, and has good selectivity when faults occur inside and outside the zone. The identification method has good adaptability to current limiting reactors with different reactance values. (3) The present invention has good speed, meets the speed requirements of longitudinal protection, and can accurately and quickly identify the fault pole when a fault occurs within the zone. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0034] Figure 1 This is a structural diagram of a four-terminal flexible DC power grid according to an embodiment of the present invention;

[0035] Figure 2 1 is a current variation trend after a fault according to an embodiment of the present invention, wherein (a) is a reverse out-of-zone fault, (b) is a forward out-of-zone fault, and (c) is an in-zone fault;

[0036] Figure 3 This is a schematic diagram of the reflection of fault waves in the zone according to an embodiment of the present invention;

[0037] Figure 4 The forward-traveling waves and the reverse-traveling waves at both ends of the fault line in the area according to the embodiment of the present invention;

[0038] Figure 5 This is a schematic diagram of the reflection of out-of-zone fault traveling waves according to an embodiment of the present invention;

[0039] Figure 6 The forward-traveling waves and the reverse-traveling waves at both ends of the out-of-zone fault line in an embodiment of the present invention;

[0040] Figure 7 This is a flow chart of a fault identification method based on line forward-reverse traveling wave correlation according to an embodiment of the present invention;

[0041] Figure 8 A voltage waveform diagram of an in-zone fault according to an embodiment of the present invention;

[0042] Fig. 9 A schematic diagram of a current fitting curve during an intra-zone fault according to an embodiment of the present invention;

[0043] Fig.10 This is a voltage waveform diagram of an out-of-zone fault according to an embodiment of the present invention;

[0044] Fig.11 A schematic diagram of a current fitting curve in an out-of-zone fault according to an embodiment of the present invention;

[0045] Fig.12 Schematic diagram of a current fitting curve during a fault outside the reverse zone according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] like Figure 1 As shown, the present invention discloses a fault identification method based on line forward-reverse traveling wave correlation, comprising:

[0049] S1, collecting real-time voltage and current signals of a multi-terminal flexible DC grid system;

[0050] S2, calculating the line mode voltage based on the real-time voltage, obtaining the line mode voltage differential value as a start criterion, returning to S1 if the line mode voltage differential value is not greater than the preset voltage value, and entering S3 if the line mode voltage differential value is greater than the preset voltage value;

[0051] S3, fitting the current signal to obtain a current change trend, and judging the fault within the zone and the forward fault outside the zone and the reverse fault outside the zone based on the current change trend;

[0052] S4. Obtain the similarity of the voltage traveling waves on the local side and the opposite side, and judge the fault within the zone and the fault outside the positive zone based on the similarity. When the fault is an internal fault, select the fault pole by using the ratio of the voltage mutation amount after the fault.

[0053] Specifically, judging the fault within the zone and the forward out-of-zone fault and the reverse out-of-zone fault based on the current change trend includes:

[0054] Fitting the current signal according to the least square method to obtain the current change trend;

[0055] An identification criterion is constructed according to the slope of the current change trend, and the in-zone fault, the forward out-zone fault, and the reverse out-zone fault are determined according to the identification criterion.

[0056] Specifically, the identification criterion is:

[0057]

[0058] Among them, S is the pole selection criterion, sgn is the sign function, and k is the slope of the current change trend.

[0059] Specifically, obtaining the similarity of the voltage traveling waves on the local side and the opposite side, and judging the fault within the zone and the fault outside the forward zone based on the similarity includes:

[0060] The reverse wave of the protection line voltage at the preset time after the fault is extracted and the forward wave of the protection line voltage at the opposite side is calculated by Pearson correlation coefficient;

[0061] The in-zone fault and the out-of-zone fault are judged according to the Pearson correlation coefficient.

[0062] Specifically, the method for judging the fault within the zone and the fault outside the positive zone according to the Pearson correlation coefficient is:

[0063]

[0064] Among them, ρ is the correlation coefficient of the voltage traveling waves on both sides of the line, ρ set is the threshold value that is set.

[0065] Specifically, when the fault is an intra-zone fault, the fault pole is selected by using the ratio of the voltage mutation amount after the fault, including:

[0066] Obtain voltage data with a time window of a preset time after the fault, and calculate the ratio of the positive and negative voltage mutations;

[0067] The faulty pole is selected based on the ratio of the voltage mutation amounts of the positive pole and the negative pole.

[0068] Specifically, the method for selecting the fault pole is:

[0069]

[0070] Among them, P set represents the setting threshold, and P is the ratio of the positive and negative voltage mutations calculated from the 2ms data after the fault.

[0071] The present invention is further described in detail below in conjunction with specific embodiments:

[0072] 1. Multi-terminal flexible DC system

[0073] This embodiment is based on the Zhangbei four-terminal flexible DC power grid. A four-terminal ±500kV flexible DC power grid system is built on the PSCAD simulation platform. The structure is as follows: Figure 1 As shown in the figure, a symmetrical bipolar connection method is adopted, and the main components include: AC power supply (S1-S4), converter station (AD), current limiting reactor (L AB , L BA , L AC , L CA , L CD , LDC , L BD , L DB ), circuit breakers (CB1-CB8), measuring devices (R AB , R BA , R AC , R CA , R BA , R CD , R DC , R BD , R DB ) and transmission lines (L1-L4), the converter station is a half-bridge MMC, converter stations A, B, and C use fixed active power control, and converter station D uses fixed DC voltage control. The current limiting reactors are all configured at the converter station exit, and the reactance values ​​are all set to 150mH. In the figure, F1-F11 are faults set at different locations. The converter station parameters and line parameters are shown in Table 1 MMC parameters and Table 2 line parameters.

[0074] Table 1

[0075]

[0076] Table 2

[0077]

[0078] 2. Identification of in-zone and out-zone faults based on Pearson correlation of fault voltage traveling waves

[0079] 2.1 Line Mode Conversion

[0080] Due to the coupling relationship between the bipolar lines, in order to avoid the influence of the coupling effect of the positive and negative lines and the accuracy of the theoretical analysis below, the Karrenbauer transformation is used to decouple them into independent mode domain networks. The transformation formula is:

[0081]

[0082] Among them, F0 and F1 represent zero-mode and line-mode voltage or current, F P With F n Represents positive and negative voltage or current.

[0083] 2.2 Direction determination of mutation current fitting based on the least squares method

[0084] This embodiment uses the method of fitting the sudden change direction of the current during the fault to judge the fault within the zone and the forward fault outside the zone and the reverse fault outside the zone. AB For the research object, Figure 2 The following are the line mode current traveling wave variation curves collected 2ms after the fault starts when the reverse fault occurs outside the zone, the fault occurs inside the zone, and the forward fault occurs outside the zone. Figure 2 (a)- Figure 2 In (c), it can be seen that the mutation trend of the reverse out-of-zone fault is negative, and the forward out-of-zone and in-zone faults are positive. AB When the current mutation direction is positive, it indicates a positive out-of-zone fault or an in-zone fault. When the current mutation direction is negative, a reverse out-of-zone fault occurs.

[0085] In order to quantify the characteristics of the fault within the zone and the forward and reverse faults outside the zone, the least square method is used to fit the direction of current change. The least square method uses the idea of ​​curve fitting to find the corresponding univariate function in the data set and obtain the form of y=kx+c. The calculation formula of k and c obtained by fitting is:

[0086]

[0087] Among them, k and c are parameters, x and y are data sets, i is the i-th data in the data set, and N is the length of the data set.

[0088] 2.3 Analysis of fault wave characteristics inside and outside the zone

[0089] When a fault actually occurs, the traveling wave will be refracted and reflected when passing through the boundary and the fault point, so it is necessary to analyze the refraction and reflection of the traveling wave when the fault occurs inside and outside the fault area.

[0090] 2.3.1 Faults within the area

[0091] The traveling wave refraction reflection diagram when an in-area fault occurs is as follows Figure 3 As shown, Figure 3 In the example, we still take the L1 line fault as an example. q1 represents the forward wave, u f1 represents the anti-travelling wave, α F With β F They represent the reflection and refraction coefficients of the traveling waves at the fault point, and the refraction and reflection coefficients at the boundaries of the converter stations at both ends are α1 and β1 respectively. Figure 3 The transmission path of the traveling wave and the reflection process of the traveling wave at the line fault point and the line boundary can be clearly seen in the figure. Since the study of the voltage traveling wave propagation path is independent of the propagation process of other line traveling waves, Figure 3 The propagation process of other lines and the refracted waves at the line boundaries are not shown.

[0092] from Figure 3 It can be seen from the figure that when an in-zone fault occurs, the line wave undergoes multiple reflection and refraction processes, which has a great influence on the change trend of its waveform. Moreover, in the same time after the protection is activated, the forward wave at the protection measurement point close to the fault point undergoes more reflection processes than the reverse wave at the protection measurement point far from the fault point. This leads to the protection R ABTerminal line mode voltage forward wave and protection R BA The similarity of the reverse traveling waves of the end line mode voltage is very low. When an in-zone fault occurs, the waveforms of the forward traveling wave at the MMCA end and the reverse traveling wave at the MMCB end are shown in Figure 4. Figure 4 The waveform in the figure is the voltage data of both ends of the line 2ms after the fault and the forward and reverse waves are extracted. Figure 4 It can be seen that the voltage waveform measured at the MMCA end is very similar to the voltage waveform measured at the MMCB end, there is almost no intersection and the shapes of the waveforms are quite different.

[0093] 2.3.2 Out-of-zone faults

[0094] When an out-of-area fault occurs on the line, the fault point is located on the left side of converter station A. The refraction and reflection process of the voltage traveling wave after the fault is as follows: Figure 5 As shown, the boundary refraction and reflection coefficients are α1 and β1 respectively. Figure 5 The vertical time axis in the middle represents the time when the traveling wave is refracted and reflected, which can clearly reflect the traveling wave components of the line mode voltage in different periods of time. Figure 5 From the transmission path of the traveling wave, it can be seen that the traveling wave of the reverse fault outside the zone occurring at the MMCA end is refracted into the L1 line through the boundary of the A end. Figure 5 The solid grey line represents the forward wave, and the dashed grey line represents the reverse wave. Figure 5 It can be seen that the forward wave and the reverse wave at the opposite end do not experience any refraction and reflection process. Therefore, except for the time difference, the line mode voltage forward wave measured at the MMCA end and the line mode voltage reverse wave measured at the MMCB end have a high degree of similarity. When an out-of-zone fault occurs, the voltage traveling wave waveforms measured at both ends are as follows: Figure 6 As shown in the figure, this is obviously different from the case of an in-zone fault. Another out-of-zone fault is a reverse fault at the MMCB end. The transmission path of this fault traveling wave is completely symmetrical with the previous one. Therefore, when a reverse fault occurs at the MMCB end, the protection R BA The measured fault voltage forward wave and protection R AB The measured fault reverse traveling waves have a high degree of similarity, so a fault identification method can be constructed based on this feature.

[0095] When a positive fault occurs, the reverse wave of the voltage on this side is extracted to judge the similarity with the forward wave on the opposite side. When a reverse fault occurs, the forward wave of the protection on this side is extracted to judge the similarity with the reverse wave of the protection on the opposite side. Therefore, the protection R AB The direction of the current mutation on the side is used as the identification criterion. The internal and external fault logic is to extract the reverse wave of the protection line mode voltage on this side and the forward wave of the voltage on the opposite side for similarity judgment when a positive direction fault occurs. When a reverse direction fault occurs, the forward wave of the voltage on this side and the reverse wave of the voltage on the opposite side are extracted for similarity judgment, so as to distinguish the internal and external faults.

[0096] 2.4 Fault Pole Selection

[0097] In the case of line faults, after determining the fault area, it is necessary to select the fault pole to determine whether it is a positive pole fault, a negative pole fault, or a bipolar short circuit fault. In this embodiment, a true bipolar four-terminal flexible DC power grid model is adopted, and the positive and negative poles are completely symmetrical structures. When no fault occurs, the voltage of the positive line is measured as +U N , the negative electrode voltage measured is -U N After a bipolar fault occurs, since the fault additional equivalent voltage is connected in series on the positive and negative lines, the voltage presents a completely symmetrical structure, and the fault additional amount of the positive and negative poles is the same, so the voltage mutation measured at the positive and negative poles is basically the same. For a single-pole grounding fault, no matter the positive or negative pole fails, the voltage change at the fault pole is greater than that at the non-fault pole, so the fault pole selection can be realized based on the ratio of the mutation amount of the positive and negative pole voltages. The formula is:

[0098]

[0099] Where: max represents the maximum value within the sampling time, u N Represents the rated voltage, u + with u － Represent the positive and negative electrode voltages respectively, and P is the mutation amount of the positive and negative electrode voltages.

[0100] 3 Identification criteria

[0101] 3.1 Startup criteria

[0102] A fast and reliable identification criterion is a prerequisite for accurate protection action. After a DC line fault occurs, the line voltage will drop rapidly. Therefore, this embodiment selects the voltage change rate as the identification start criterion. In order to avoid the coupling effect between lines, the line mode voltage change rate is selected as the fault start criterion. The criterion expression is:

[0103] Δu1=|u1(t)-u1(t-1)|≥Δu 1set

[0104] Where: u1 is the line mode voltage, Δu1 is the difference between adjacent line mode voltages. 1set The threshold value should be set to avoid the maximum line mode voltage change rate when the line is in steady state operation, and the sensitivity and reliability of identification should be considered comprehensively. The threshold value is set to 200kV / ms. Among them, the line mode voltage is calculated by the Karrenbauer transformation formula

[0105] 3.2 Identification Criteria

[0106] In the current change curve, it can be observed that when a reverse out-of-zone fault occurs, the current mutation trend is negative, while when an in-zone fault and a forward out-of-zone fault occur, the current change direction is positive, so this can be used as an identification criterion, and either side of the line is taken as the object of protection direction judgment. After fitting the current change trend k value (k value is the slope), it is only necessary to determine the positive or negative of k to determine the in-zone fault, the forward out-of-zone fault, and the reverse out-of-zone fault, so the sign function sgn is introduced to construct the criterion S. In order to avoid the interference of abnormal data, the length of the data window is selected as 2ms for line mode current traveling wave fitting, and the value of k is fitted. If k is greater than 0, S=sgn(k) is 1, and it is judged that an in-zone fault and a forward out-of-zone fault occur. If k is less than 0, S=sgn(k) is -1, and it is judged that a reverse out-of-zone fault occurs. The identification and judgment formula is:

[0107]

[0108] 3.3 Fault area identification

[0109] The Pearson correlation theory is used to compare the similarity of the voltage traveling waves at both ends. The Pearson correlation is an indicator for comparing the similarity of two time series. The calculation formula of the correlation coefficient is:

[0110]

[0111] Where: X, Y represent two time series, n is the number of data elements in the sequence, ρ X,Y Represents the correlation coefficient between the X series and the Y series.

[0112] The Pearson correlation coefficient is only related to the trend of the data set, and will not change due to the change of the scale and position of the two variables. Therefore, when measuring the similarity of the voltage traveling wave, it will not change due to the difference in the amplitude of the voltage traveling wave at both ends, but only related to the linear relationship between the data sets. The smaller the value of the calculated correlation coefficient, the weaker the correlation between the two data sets. The data range in Table 3 represents the degree of correlation between the two data sets. Compared with the similarity calculation methods based on the principles of Euclidean distance and DTW distance, the advantage of the Pearson correlation coefficient is that it does not require standardization operations due to the large difference in the amplitude of the two sequences; compared with the cosine similarity algorithm, the types of these two algorithms are more similar, and in many cases the two can be regarded as equivalent, but the Pearson correlation coefficient centers the signal, which can cope with the situation of missing signal acquisition, and has a certain inhibitory effect on noise and zero drift, and is more inclined to the results of steady-state data.

[0113] Table 3

[0114]

[0115] After a fault occurs, the protection RAB Determine the fault direction. If S=1, the fault direction is judged to be in the positive direction. The reverse wave of the protection on this side 2ms after the fault and the forward wave of the protection on the opposite side are extracted for correlation calculation. If S=-1, the fault direction is judged to be in the reverse direction. The forward wave of the protection on this side 2ms after the fault and the reverse wave of the protection on the opposite side are extracted for correlation calculation. After a large number of simulation calculations and the above analysis, it can be deduced that when an out-of-zone fault occurs, the similarity of the voltage traveling waves on both sides is close to 1, and when an in-zone fault occurs, the correlation coefficient of the traveling waves on both sides is close to 0. Therefore, the identification criteria for inside and outside the zone are as follows:

[0116]

[0117] Where: ρ is the correlation coefficient of the voltage traveling waves on both sides of the line, ρ set is the set threshold. After a large number of simulations, it is found that when the fault occurs outside the zone, the calculated correlation coefficient is greater than 0.9, while when the fault occurs inside the zone, the correlation coefficient is less than 0.5. In order to correctly identify the fault, the setting threshold is 0.8.

[0118] 3.4 Fault selection criteria

[0119] When it is determined that the fault occurs in the zone, this embodiment uses the ratio of the voltage mutation amount after the fault to select the fault pole. If a positive pole fault occurs, the ratio of the positive pole to the negative pole voltage mutation amount is greater than P set When a negative electrode fault occurs, the ratio of the mutation amount is less than 1 / P set When a bipolar short circuit fault occurs, the ratio of the mutation amount is greater than 1 / P set and less than P set , so the fault selection criterion can be obtained as:

[0120]

[0121] Where: P set represents the threshold value of setting, and P is the ratio of the positive and negative voltage mutations calculated by taking the time window as 2ms after the fault. set The value is set to 2.

[0122] 3.5 Fault Identification Process

[0123] Based on the above analysis, the design Figure 7 The fault identification process shown in the figure calculates the line mode voltage and line mode current by real-time acquisition of voltage and current. If the line mode voltage change rate is greater than the setting threshold, it is judged that the fault identification is started after the fault occurs; the fault data time window is selected as 2ms data after the fault, and the protection R is calculated by the least squares method. ABThe extracted line mode current fitting k value is used to judge the fault within the zone and the forward fault outside the zone and the reverse fault outside the zone. The line mode voltage forward wave and reverse wave at both ends of the line 2ms after the fault is started are extracted, and their Pearson correlation coefficient is calculated to determine the fault area; then the ratio of the positive and negative voltage mutations is calculated by the positive and negative voltages 2ms after the fault to select the fault pole.

[0124] 4. Validity Verification

[0125] In order to verify the effectiveness of the proposed identification scheme, this embodiment builds a four-terminal simulation model on the PSCAD simulation platform, and sets the reverse direction out-of-zone fault locations as the converter station A exit, S1 exit bus, and L2 line out-of-zone fault in the model, and the forward direction out-of-zone fault locations as the converter station B exit, S2 exit bus, and L3 line out-of-zone fault.

[0126] 4.1 In-zone fault simulation verification

[0127] Assumptions Figure 1 A positive ground fault occurs at F3, the transition resistance is set to 0.01Ω, and the fault location is 100 meters from the protection R AB The side head end is 105km, and the protection R is calculated by the least squares method. AB The side k value is 4.75×10 7 , c value 1×10 4 , and the fitting curve is obtained as Fig. 9 As shown, S=1 is calculated, and it is judged to be a positive direction fault, such as Figure 8 The voltage waveform diagram is shown in the figure when the fault occurs in the area. The protection R AB The voltage reverse wave on the side and the protection R BA The voltage traveling wave curve on the two sides is compared and the Pearson correlation coefficient ρ of the two curves is calculated. The shapes of the voltage traveling wave curves at both ends are different. The value of ρ is calculated to be 0.0237, so it is determined to be an internal fault.

[0128] 4.2 Out-of-zone fault simulation verification

[0129] 4.2.1 Fault outside the forward zone

[0130] Assumptions Figure 1 The four-terminal flexible DC grid shown in FIG. 1 has a bipolar short-circuit fault outside the positive region with a transition resistance of 0.01Ω. Figure 1 Taking fault F6 as an example, the k value is calculated by the least square method to be 3.07×10 7 The c value is 1.74×10 3 , and the fitting curve is obtained as Fig.11 As shown, S=1 is calculated, which is judged as a positive direction fault. At this time, R AB The reverse wave on the side and R BA The side-to-side wave curve is as follows Fig.10 As shown in Figure 1, the Pearson correlation coefficient is used to calculate the traveling wave correlation coefficient ρ at both ends, which is 0.9723, so it is determined to be a fault outside the forward zone.

[0131] 4.2.2 Fault outside the reverse zone

[0132] Assumptions Figure 1 The four-terminal MTDC power grid shown in the figure has a negative short-circuit fault outside the reverse direction zone with a transition resistance of 0.01Ω. The fault location is as follows Figure 1 As shown in F7 in Figure 1, the value of k is -1.49×10 5 The value of c is -284.63, and the fitting curve is as follows Fig.12 As shown, S=-1 is calculated, and it is determined to be a reverse direction out-of-zone fault.

[0133] 4.3 Verification of fault identification methods with different transition resistances

[0134] When a fault actually occurs, a short-circuit fault is often not metallic, but must experience a certain degree of transition resistance. In order to verify whether the identification method proposed in this embodiment can work correctly under transition resistance, faults are set at different positions on the L1 line, and the fault resistance is set to 0.01Ω, 100Ω, and 300Ω respectively. The simulation results under different transition resistances are shown in Table 4, and the pole selection results under different transition resistances for intra-zone faults are shown in Table 5. The simulation verification shows that the identification method proposed in this embodiment can still accurately distinguish the fault area for faults occurring at different positions under different transition resistances, and can accurately select the fault pole for intra-zone faults.

[0135] Table 4

[0136]

[0137] Table 5

[0138]

[0139] 4.4 Effect of current limiting reactor on fault identification

[0140] The current limiting reactor has always been an indispensable part of the flexible DC line, and the size of the current limiting reactor has always been an important factor affecting the effectiveness of the identification method. In order to explore whether the identification method is still applicable under current limiting reactors with different amplitudes, the amplitude of the current limiting reactor is set to 75mh, 150mh, and 200mh in this embodiment. The simulation results under different current limiting reactors are shown in Table 6. It can be seen from the simulation results in Table 6 that for current limiting reactors of different sizes, the proposed identification method has good applicability.

[0141] Table 6

[0142]

[0143] 4.5 Effect of noise on fault identification

[0144] The protection device receives information in the form of receiving signals. The process of signal propagation will inevitably be interfered by noise. In view of the fact that the current protection measurement device is generally subject to noise with a signal-to-noise ratio greater than or equal to 30db, Gaussian white noise with a signal-to-noise ratio of 30db is mixed into it to verify the effectiveness of the recognition method under noise interference. The simulation results before and after the noise is mixed are shown in Table 7. It can be seen from the results in Table 7 that after the noise is mixed, the recognition method proposed in this embodiment can still operate correctly.

[0145] Table 7

[0146]

[0147] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A fault identification method based on line forward-reverse wave correlation, characterized in that: include: S1, collecting real-time voltage and current signals of a multi-terminal flexible DC grid system; S2, calculating the line mode voltage based on the real-time voltage, obtaining the line mode voltage differential value as a start criterion, returning to S1 if the line mode voltage differential value is not greater than the preset voltage value, and entering S3 if the line mode voltage differential value is greater than the preset voltage value; S3, fitting the current signal to obtain a current change trend, and judging the fault within the zone and the forward fault outside the zone and the reverse fault outside the zone based on the current change trend; Judging the fault within the zone, the forward fault outside the zone, and the reverse fault outside the zone based on the current change trend includes: Fitting the current signal according to the least square method to obtain the current change trend; Establishing an identification criterion according to the slope of the current change trend, and judging the in-zone fault, the forward out-zone fault, and the reverse out-zone fault according to the identification criterion; The identification criteria are: ； Among them, S is the pole selection criterion, sgn is the sign function, and k is the slope of the current change trend; S4. Obtain the similarity of the voltage traveling waves on the local side and the opposite side, and judge the fault within the zone and the fault outside the positive zone based on the similarity. When the fault is an internal fault, select the fault pole by using the ratio of the voltage mutation amount after the fault.

2. The fault identification method based on line forward-reverse traveling wave correlation according to claim 1 is characterized in that: Acquiring the similarity of the voltage traveling waves on the local side and the opposite side, and judging the fault within the zone and the fault outside the forward zone based on the similarity includes: The reverse wave of the protection line voltage at the preset time after the fault is extracted and the forward wave of the protection line voltage at the opposite side is calculated by Pearson correlation coefficient; The in-zone fault and the out-of-zone fault are judged according to the Pearson correlation coefficient.

3. The fault identification method based on line forward-reverse traveling wave correlation according to claim 2 is characterized in that: The method for judging the fault within the zone and the fault outside the positive zone according to the Pearson correlation coefficient is: ； Among them, ρ is the correlation coefficient of the voltage traveling waves on both sides of the line, ρ set is the threshold value that is set.

4. The fault identification method based on line forward-reverse wave correlation according to claim 1 is characterized in that: When the fault is an internal fault, the fault pole is selected by using the ratio of the voltage mutation after the fault, including: Obtain voltage data with a time window of a preset time after the fault, and calculate the ratio of the positive and negative voltage mutations; The faulty pole is selected based on the ratio of the voltage mutation amounts of the positive pole and the negative pole.

5. The fault identification method based on line forward-reverse wave correlation according to claim 4 is characterized in that: The method for selecting the fault pole is: ； Among them, P set represents the setting threshold, and P is the ratio of the positive and negative voltage mutations calculated from the 2ms data after the fault.

Images