CN109888744B

CN109888744B - Protection method for high-voltage direct-current transmission line

Info

Publication number: CN109888744B
Application number: CN201910159693.1A
Authority: CN
Inventors: 戴志辉; 刘宁宁; 张程
Original assignee: North China Electric Power University
Current assignee: North China Electric Power University
Priority date: 2019-03-04
Filing date: 2019-03-04
Publication date: 2021-01-12
Anticipated expiration: 2039-03-04
Also published as: CN109888744A

Abstract

The invention discloses a protection method of a high-voltage direct-current transmission line, which comprises the following steps: step 1: determining criteria for identifying faults inside and outside the area; step 2: determining a structural locking criterion; and step 3: determining a fault starting criterion; and 4, step 4: determining fault pole selection criteria; and 5: and determining a protection scheme book.

Description

Protection method for high-voltage direct-current transmission line

Technical Field

The invention relates to the technical field of circuit protection, in particular to a protection method of a high-voltage direct-current transmission line.

Background

High Voltage Direct Current (HVDC) is rapidly developed and widely applied to long-distance large-capacity power transmission and asynchronous grid interconnection. The direct-current transmission line has long distance, complex crossing area and high fault occurrence rate, and the reliable and perfect direct-current line protection is more important. At present, in the operated direct current engineering, traveling wave protection and differential undervoltage protection are mostly adopted for main line protection, and low-voltage protection and current differential protection are configured for backup protection. Traveling wave protection needs to identify a traveling wave head, the wave head is difficult to detect when a high-resistance grounding fault occurs, and the problems of high sampling rate, complex threshold value setting and the like exist; differential under-voltage protection utilizes line voltage differential and amplitude to form a protection criterion, and has poor transition resistance capability; the pilot current differential protection is mainly used for identifying the high-resistance grounding fault of the direct-current line, but the existing differential protection only utilizes current data at two ends to simply construct identification criteria, does not consider the influence of capacitance current in the fault transient process of a long-distance large-capacity direct-current transmission line, and has slow action speed which is even as long as 1.1 s.

The existing high-voltage direct-current line protection is divided into double-end quantity protection and single-end quantity protection. At present, the research focus of double-end line protection is mainly based on the traveling wave principle and mutation analysis. The need to transmit waveform data places high demands on the communication channel. Single-ended line protection is mainly based on the boundary characteristics of the dc system. However, many methods neglect the attenuation effect of the power transmission line on the high-frequency signal, cannot protect the full length of the line, and have complex matching relationship.

Therefore, a protection method for the high-voltage direct-current transmission line is expected to solve the problems that in the prior art, the pilot current differential protection of the high-voltage direct-current transmission line is poor in rapidity and limited in transition resistance tolerance.

Disclosure of Invention

The invention discloses a protection method of a high-voltage direct-current transmission line, which is characterized by comprising the following steps of:

step 1: determining criteria for identifying faults inside and outside the area; when the direct current transmission line is in an internal fault, the identification factors at the two ends are used for judging the fault, and when the rectifying side is in an external fault, the identification factor X at the rectifying end is smaller than the identification criterion threshold value k and is directly judged as the external fault; when the inversion side out-of-area fault occurs, the inversion end identification factor Y is smaller than a criterion threshold value k and is directly judged as the out-of-area fault;

step 2: determining a structural locking criterion;

and step 3: determining a fault starting criterion;

and 4, step 4: determining fault pole selection criteria;

and 5: determining a protection scheme, continuously collecting current and voltage data on two sides of a line boundary, extracting specific frequency band components, starting a protection starting element at one end if the voltage of a protection installation position of a direct-current line at the end meets a protection starting criterion, judging whether the side protection is locked or not according to the locking criterion, and calculating the transient energy value of the specific frequency band of the corresponding data window time length on the two sides of the end boundary to obtain an identification factor of the end if the protection is not locked; when the identification factor of any end of the direct current line is smaller than a threshold value k, judging as an out-of-area fault by using single-end fault information; if the identification factors X, Y of the rectifying end and the inverting end are both larger than a threshold value k, determining that the fault is in the area; and if the fault is judged to be in the region, calculating a pole selection function W by using the voltage fault component.

Preferably, in step 1, the current and voltage fault components of a specific frequency band are extracted by using a chebyshev filter, and both the current and voltage fault components are extracted within a time period from the fault to the completion of the adjustment of the control system.

Preferably, step 1 defines the i-order transient energy values of the specific frequency bands at the measurement points m, n, a, b as:

wherein:

in the formula, p_m-fil、p_n-fil、p_a-fil、p_b-filRespectively sampling values of 2.5-5 kHz frequency band transient power at the measuring points m, n, a and b, wherein the values are equal to corresponding current and voltage fault components i_m-fil、i_n-fil、i_a-fil、i_b-fil、u_m-fil、u_n-fil、u_a-fil、u_b-filThe product of the two; j ═ 1, 2, …, J; j is the number of sampling points within 5 ms; i represents the order of the transient energy value, E when i is 4⁽⁴⁾ _m-filRepresenting the 4-order transient energy of m point 2.5-5 kHz frequency band, and the value of the energy is equal to p in the data window_m-filThe sum of the absolute values of the fourth power;

the identification factors X, Y at the ends of the rectifying side and the inverting side are defined as follows:

constructing the internal and external fault identification criteria as follows:

when the identification factors X, Y at the rectifying side and the inverting side are both greater than the threshold value k, the fault is judged to be an intra-area fault; when any party in X, Y is less than k, it is judged as an out-of-range fault;

according to the formulas (15) and (16), P is a fault in the high frequency band_m/P_a>>1，P_n/P_b>>1; in case of an external fault in the rectifying side zone, P_m/P_a<<1; when an external fault occurs on the inversion side, P_n/P_b<<1。

Preferably, the construct latch-up criterion of step 2 is based on the waveform match error K shown in equation (17)_pTo measure the difference in current waveform:

wherein x and y may be m, a or n, b, each independently representing i_m-filAnd i_a-filOr i_n-fil、i_b-filThe constructed locking criterion is as follows (18):

in the formula (18), K_setTaking K as the minimum value of the current waveform matching error on both sides of the boundary after the occurrence of the internal and external faults in the normal operation time zone of the direct current filter for the threshold value of the locking criterion_set＝0.2。

Preferably, the fault starting criterion of step 3 is as shown in formula (19):

in the formula, v_m-fil、v_n-filSampling values of 2.5-5 kHz voltage at the m and n measuring points respectively; j ═ 1, 2, …, J; j is the number of sampling points within 5 ms; v. of_setFor threshold values, v, set in the starting criterion_set＝1kV。

Preferably, said step 4 determines the pole selection criterion of formula (21) according to the constructed pole selection function of formula (20):

in the formula,. DELTA.v_m1、Δv_m2Voltage fault components measured at the line protection installation positions of the positive rectifying side and the negative rectifying side are respectively, and the value of the voltage fault components is equal to the difference between transient voltage after fault and corresponding voltage in normal operation; w_set1、W_set2Threshold value, W, of pole selection criterion for faults_set1＝1.5、W_set2＝0.6。

The method comprises the steps of firstly analyzing impedance frequency characteristics of line boundaries, and researching additional networks for fault components inside and outside a direct current transmission line by combining a superposition principle to find that when a fault occurs inside a region, the energy ratio of specific frequency bands on two sides of the boundary at any end of two ends of the line is a number far larger than 1; when the rectifying end is out of area, the specific frequency band energy ratio at two sides of the line boundary at the rectifying end is far less than 1; when the inversion end has an external fault, the energy ratio of the specific frequency bands at two sides of the line boundary of the inversion end is far less than 1. Thus, a protection scheme based on the specific band energy ratio on both sides of the line boundary is proposed. And finally, an upward extra-high voltage direct current transmission engineering model is built in the PSCAD/EMTDC, and the correctness of the protection scheme is verified through simulation. The result shows that the protection principle is simple, the calculated amount is small, the full length of the line can be reliably protected, the transition resistance capability is good, and the rapidity is superior to that of the conventional differential current protection.

Drawings

Fig. 1 is a schematic diagram of the basic structure of a bipolar HVDC system.

Fig. 2 is a schematic diagram of a filtering link formed by an upward engineering direct current filter and a smoothing reactor.

Fig. 3 is a graph of the impedance-frequency characteristics of the upper engineered dc filter.

Fig. 4 is a schematic diagram of a fail-over network of the post-zone fail-over system.

Fig. 5 is a schematic diagram of a fail-over network for a post-zone system.

FIG. 6 shows the upward engineering Z_lb/(Z_p+Z_lb) Impedance frequency characteristic graph of (1).

Fig. 7 is a flow chart of a protection scheme.

FIG. 8 shows the starting voltage v_m-fil、v_n-filAnd absolute value mean curve diagram thereof.

Fig. 9 is a graph of specific band current on both sides of a boundary in the event of an intra-zone fault.

Fig. 10 is a graph of voltage of a specific frequency band on both sides of a boundary at the time of an intra-zone fault.

Fig. 11 is a graph of transient power for specific frequency bands on both sides of a boundary in the event of an intra-zone fault.

FIG. 12 is the fail select voltage Δ v_m1、Δv_m2A graph of (a).

Fig. 13 is a graph of the high frequency component of the voltage experienced at the line head end protection installation.

FIG. 14 shows the starting voltage v_m-fil、v_n-filAnd absolute value mean curve diagram thereof.

Fig. 15 is a graph of current, voltage, transient power for specific frequency bands across an out-of-band fault boundary.

Detailed Description

In order to make the implementation objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be described in more detail below with reference to the accompanying drawings in the embodiments of the present invention. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are only some, but not all embodiments of the invention. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The high-voltage direct-current transmission system comprises: the basic structure of the bipolar HVDC system shown in fig. 1 mainly comprises a rectifier station, an inverter station and a dc transmission line 3. Smoothing reactor L_pAnd the direct current filter form a direct current line boundary; a. b, m and n are measuring points on the boundary valve side and the line side of the two ends of the line respectively; f. of₁～f₅Indicating the location of the fault occurrence: f. of₁Indicating a fault point within the DC line, f₂Indicating a fault point outside the smoothing reactor on the rectifying side, f₃Indicating a fault point outside the smoothing reactor on the inverter side, f₄Indicating a fault point on the AC busbar of the rectifier side, f₅Indicating a fault point on the inverter side ac bus.

And D, direct current filtering: the filtering link of the direct current transmission line consists of a smoothing reactor and a direct current filter to form a direct current system line boundary. The direct current filtering link is explained by taking a +/-800 kV upward ultrahigh voltage direct current transmission demonstration project as an example.

2/12/39 three-tuning direct current filters are arranged on each of the positive pole and the negative pole of the two ends of the direct current line of the upward engineering, 2 smoothing reactors of 75mH are arranged on the polar line, and fig. 2 is a schematic diagram of the direct current filtering link of the rectifying end and the positions of measuring points on the two sides. In the figure: a is the installation position of the boundary valve side current divider and the voltage divider, m is the installation position of the boundary line side current divider and the voltage divider, R, C₁、L₁、C₂、L₂、C₃、L₃All parameters of the direct current filter are parameters, and specific values are shown in table 1.

TABLE 1 upward engineering DC Filter parameters

The upward engineered dc filter impedance versus frequency characteristic as shown in fig. 3 is plotted according to the parameters of table 1. As can be seen from fig. 3: when the frequency is less than 100Hz, the direct current filter has a capacitive characteristic; at the harmonic frequencies (600Hz and 1950Hz), the dc filter impedance is approximately 0; in the high frequency band of 2.5kHz to 40kHz, the DC filter can be equivalent to a pure inductor with an inductance value of about 8 mH.

And (3) analyzing the faults in the area: according to the superposition principle, after the direct current line has a fault, the fault state of the system can be equivalent to the superposition of the normal operation state and the fault additional state, and at the moment, the fault additional network is as shown in fig. 4 (the invention takes the direction from the bus to the line as the positive direction).

As shown in FIG. 4, I_m、I_n、U_m、U_nThe current and the voltage of the fault component measured by the current divider and the voltage divider at m and n respectively (the values of the current and the voltage are equal to the difference between the transient current and the voltage after the fault and the corresponding current and voltage during normal operation); i is_a、I_b、U_aAnd U_bFault component current and voltage measured by the shunt and the voltage divider at a and b respectively; u shape_F|0|Is an additional voltage source with a value equal to the fault point in normal operationThe voltage value of F; r_FIs a transition resistance; z_smAnd Z_snThe equivalent impedance of the current converter at two ends; z_pIs equivalent impedance (Z) of smoothing reactor_p＝jωL_p)；Z_lbIs the equivalent impedance of the DC filter bank; (Z)_cAnd gamma) represents a distributed parameter transmission line; z_x、Z_yThe equivalent impedances of the transmission line from the fault point to the rectifying end and the inverting end are respectively.

The rectifying end is taken as an example for detailed description. As can be seen from fig. 4, when there is an internal fault, the current and voltage fault components on both sides of the rectifying edge boundary have the following relationships:

when the current or voltage frequency is high (more than 1kHz), the equivalent impedance Z of the converters at the rectifying end and the inverting end_sm、Z_snAll containing only an electric resistance component, i.e. Z_sm、Z_snCan be respectively expressed as:

in the formula L_sm、L_smIs the equivalent internal inductance of the current converter at two ends, and omega is the angular frequency. The DC filter can be equivalent to a fixed inductor L in a high frequency band (such as 2.5-40kHz)_lbThen, the formulas (1) and (2) can be converted into formulas (4) and (5):

it is easy to know the ratio I of high-frequency fault components of current and voltage at two sides of boundary when fault occurs in region_m/I_a、U_m/U_aAre all real numbers greater than 1. The following results were obtained: in high frequency, I_m/I_a>>1 (value of about 20-40, the ratio is larger if the equivalent internal inductance of the inverter is considered). Taking the Pueraria DC transmission project as an example, the inductance of the smoothing reactor is 300mH, the equivalent inductance in the high-frequency band of the DC filter is about 18mH, the equivalent inductance of the converter is 233.8mH, and then I_m/I_a≈30.66，U_m/U_a≈2.28。

Similarly, for the inverter side, the high-frequency fault components of the current and the voltage on the two sides of the boundary of the inverter side also have a similar relation, as shown in formula (6):

the transient power of points a, b, m and n in fig. 4 is defined as follows:

in the case of the combination of the formulae (4), (5) and (6) and the above analysis, it is easy to know that P is present_m/P_a、P_n/P_bAre all numbers far greater than 1 (values greater than 20), i.e.

In summary, when an intra-zone fault occurs, the high-frequency transient power on the boundary line side (i.e., points m and n) is much larger than the power value on the boundary valve side (i.e., points a and b).

Analyzing an out-of-range fault: after the rectifying side out-of-zone fault occurs, the fault attached network is as shown in fig. 5.

As can be seen from fig. 5, the current and voltage fault components on both sides of the line boundary at the rectifying end (fault end) have the following relationships:

as shown in fig. 5, after the rectifying-side out-of-range fault occurs, the high-frequency signal generated at fault point F is transmitted to point m by the attenuation of the line boundary, theoretically I_m/I_a、U_m/U_aBoth numbers are smaller than 1, and in order to obtain the maximum value that may exist and at the same time facilitate the analysis and calculation, the direct current lines are regarded as short circuit and open circuit in equations (9) and (10), respectively, and then equations (9) and (10) can be converted into:

similar to the in-zone fault analysis, in the high frequency band, equation (11) can be converted into:

in a high frequency band (such as 2.5-40kHz), the equivalent inductance L of the DC filter_lbFar smaller than smoothing reactor inductance L_pThen at this time I_m/I_a≈0.5，U_m/U_aApproximately equal to 0 (value of approximately 0.025-0.05), and further P_m/P_a<<1 (value of about 0.0125 to 0.025).

To more intuitively derive the above theory, FIG. 6 shows Z in the upward engineering_lb/(Z_p+Z_lb) Impedance frequency characteristic curve of (1). As is apparent from fig. 6: when the frequency is less than 100Hz, the amplitude is equal to 1; at the tuning frequency, the impedance amplitude is small, approximately 0; when the frequency is 2.5-40kHz, Z_lb/(Z_p+Z_lb) Is close to 0 (about 0.05). This is consistent with theoretical analysis: for direct current and lower frequency componentsThe DC filter can be approximated as an open circuit with a large impedance value, while the smoothing reactor can be approximated with a 0 impedance value, so that Z_lb/(Z_p+Z_lb) Approximately equal to 1; for tuned frequencies, the DC filter can be approximated as a short circuit, with the smoothing reactor impedance at a non-zero value, and thus Z_lb/(Z_p+Z_lb) Is approximately 0; in the high frequency band (2.5-40 kHz), the DC filter is equivalent to a pure inductor and its inductance value L_lbFar smaller than smoothing reactor inductance L_pSo that at this time U_m/U_a＝Z_lb/(Z_p+Z_lb)≈0。

And (3) finding the conditions of faults inside and outside the comparison area: in the event of an in-zone fault, in the high-frequency band, P_m/P_aAnd P_n/P_bReal numbers (values greater than 20) that are all much greater than 1; in case of an external fault in the rectifying side zone, P_m/P_aA number much less than 1 (about 0.0125 to 0.025); when an external fault occurs on the inversion side, P_n/P_bIs a number much less than 1 (with a value of about 0.0125 to 0.025), and thus a fault identification criterion can be constructed.

And (3) identifying criteria of faults inside and outside the area: first, a chebyshev filter is used to extract a current and voltage fault component of a specific frequency band. Selecting a high-frequency band within the range of 2.5-40 kHz; considering the characteristics of low content of high-frequency components, easiness in lightning stroke interference and high requirement on sampling frequency, the selection frequency is not too high, and current and voltage fault components in a frequency band of 2.5-5 kHz are selected in the embodiment.

The current and voltage fault components are all in the time from the fault to the time before the control system is adjusted, and research shows that the process needs at least 5 ms; the high-frequency signal utilized by the protection criterion of the embodiment has a fast attenuation speed, so that the transient characteristic has short existence time, and therefore, the length of the data window is not suitable to be too long; meanwhile, the duration of the lightning interference is generally 3ms, and the time length of the finally selected data window is 5 ms.

Defining i-order transient energy values of specific frequency bands of m, n, a and b as:

wherein:

in the formula, p_m-fil、p_n-fil、p_a-fil、p_b-filRespectively sampling values of 2.5-5 kHz frequency band transient power at the measuring points m, n, a and b, wherein the magnitudes of the sampling values are equal to corresponding current and voltage fault components i_m-fil、i_n-fil、i_a-fil、i_b-fil、u_m-fil、u_n-fil、u_a-fil、u_b-filThe product of the two; j ═ 1, 2, …, J; j is the number of sampling points within 5 ms; i represents the order of the transient energy value, and takes an integer not less than 1, if i is 4, then E⁽⁴⁾ _m-filRepresenting the 4-order transient energy of m point 2.5-5 kHz frequency band, and the value of the energy is equal to p in the data window_m-filThe sum of the absolute values of the fourth power.

when the identification factors X, Y at the rectifying side and the inverting side are both greater than the threshold value k, the fault is judged to be an intra-area fault; when either side X, Y is less than k, it is judged as an out-of-range fault.

In the high frequency band, P is observed when faults in the regions (15) and (16) occur_m/P_a>>1，P_n/P_b>>1; in case of an external fault in the rectifying side zone, P_m/P_a<<1; when an external fault occurs on the inversion side, P_n/P_b<<1, then it is easy to know: the higher the order i of the transient energy E is, the more obvious the difference of the internal and external fault characteristics is, and the threshold value is easy to set; meanwhile, the increase of i is beneficial to eliminating the influence of high-frequency quantity fast attenuation on the identification criterion, namely, a higher order can be taken in practice to fully ensure the reliability of the protection action. In this embodiment, a simulation experiment chart with an order i of 4 is given, a certain margin is reserved in consideration of the influence of measurement errors, and a threshold value k may be set to 1.

When a fault occurs in a direct current transmission line area, identification factors at two ends are needed to judge the fault, but only state signals are needed to be transmitted, waveform data and the like do not need to be transmitted in real time, and therefore the requirement on synchronism is low.

When the rectifying side has an external fault, the identification factor X is smaller than the identification criterion threshold value, and the external fault can be directly judged without inverting side fault data; when the outside fault of the inversion side is detected, the identification factor Y is smaller than a criterion threshold value, and the outside fault can be directly judged without rectifying side data.

It is worth noting that when the direct current filter exits operation due to fault, the situations of operation rejection, misoperation and the like may occur, and a locking criterion is added for improving the protection reliability.

Locking criterion: when the direct current filter quits operation due to faults, currents on two sides of the boundary are almost the same, the waveforms are basically consistent, and the matching degree is high; when the direct current filter is normally switched in, the high-frequency current waveforms on the two sides of the fault rear boundary have obvious difference and the matching degree is low.

Using the waveform matching error K shown in equation (17)_pDifference of measurement current waveform:

wherein x and y may be m, a or n, b, each independently representing i_m-filAnd i_a-filOr i_n-fil、i_b-filThe waveform matching error of (1). The criteria for constructing the latch are as follows:

in the formula (18), K_setThe method is characterized in that a threshold value of a locking criterion is adopted, when the direct current filter is out of operation, current waveforms on two sides of a boundary are basically matched, and a matching error is close to 0; when the direct current filter is normally put in, the high-frequency current waveforms on two sides of the fault rear boundary have large difference and large matching error, a certain margin is taken in consideration of the measurement error, and K is finally set_set＝0.2。

Fault starting criterion: when the device normally operates, the high-frequency band voltages at the measuring points m and n are both approximate to zero; after the fault occurs, the voltages of the high-frequency sections at m and n are obviously increased. Thus, the following protection start criteria are constructed:

in the formula, v_m-fil、v_n-filSampling values of 2.5-5 kHz voltage at the m and n measuring points respectively; j ═ 1, 2, …, J; j is the number of sampling points within 5 ms; v. of_setIs a threshold value set in the start criterion. Because the high-frequency signal is attenuated after being transmitted by the line boundary and the long line, and the transition resistance also reduces the amplitude of the high-frequency component in the fault, the setting of the threshold value needs to consider the voltage quantity of the inversion side (or rectification side) in the fault outside the rectification side (or inversion side) and the voltage quantity of the rectification side (or inversion side) in the fault of the high-resistance grounding at the tail end (or head end) of the line_set＝1kV。

Fault pole selection criterion: for a bipolar direct-current power transmission system, when a single-pole fault occurs, the amplitude of a fault component of a fault pole voltage is larger than that of a non-fault pole; when the bipolar fault occurs, the amplitude of the voltage fault component of the two poles is close, and therefore, the pole selection function is constructed as follows:

the pole selection criterion is as follows:

in the formula,. DELTA.v_m1、Δv_m2Voltage fault components (the value of which is equal to the difference between transient voltage after fault and corresponding voltage in normal operation) measured at the line protection installation positions of the positive rectifying side and the negative rectifying side respectively; w_set1、W_set2And selecting a threshold value of the criterion for the fault pole selection. The coupling coefficient of two-pole DC transmission line on the same tower is generally less than 0.5^[3,17-18](ii) a Considering that when the tail end of the line has a high-resistance earth fault, the voltage signal measured by the rectifying side of the fault electrode possibly has little difference with the voltage signal strength measured by the rectifying side of the non-fault electrode under the attenuation action of the long line, and considering a certain margin, selecting W_set1＝1.5、W_set2＝0.6。

The protection logic is that both sides of the line boundary continuously collect current and voltage data (and extract specific frequency band components thereof), if the voltage at the protection installation position of a direct current line at a certain end meets a formula (19), the protection starting element at the end is started, whether the side protection is locked is judged according to a formula (18), if the protection is not locked, the transient energy value of the specific frequency band of the corresponding data window time length (5ms) at both sides of the end boundary is calculated according to a formula (13), and the identification factor of the end is obtained by using a formula (15); when the identification factor of any end of the direct current line is smaller than a threshold value k, the out-of-area fault can be judged by only utilizing single-end fault information; if the identification factors X, Y of the rectifying end and the inverting end are both larger than a threshold value k, determining that the fault is in the area; if it is determined that there is an intra-area fault, a selection function W is calculated according to equation (20) using the voltage fault component, and selection of a fault pole is realized according to equation (21). Fig. 7 is a flow chart of the protection scheme.

A simulation model of the +/-800 kV upward high-voltage direct-current transmission project shown in the figure 1 is established by utilizing PSCAD/EMTDC electromagnetic transient simulation software. The rated transmission power of the direct current system is 6400 MW, the total length of the line is 1891 km, two-pole direct current lines are connected on the same pole, the lines adopt a frequency-dependent parameter model and are all 6 split conductors, the inductance value of a smoothing reactor is 150mH, and 2/12/39 triple-tuned direct current filters are arranged at two ends of the lines (the wiring structure and parameters of a direct current filtering link are respectively shown in figure 2 and table 1). The fault occurs at the time t-3 s, and the fault duration is 10 s. The sampling frequency was set to 20 kHz. Space limitation, only a simulation experiment chart based on the 4-order transient energy ratio is given.

Simulation results of in-zone faults: when the midpoint of the positive transmission line is in ground fault through the 50 omega transition resistor (the direct current filter is normally put into use), the protection characteristic quantity can be seen from the graph of fig. 8, and after the fault occurs, the voltages v of the specific frequency bands at the m and n ends of the direct current transmission line_m-fil、v_n-filAre all obviously increased: v is within 3.000-3.005 s_m-fil、v_n-filAre all larger than a starting criterion threshold value v_setAnd (5) starting protection at both ends of m and n under the condition of 1 kV. As shown in fig. 9, the waveform difference of the specific frequency band current at two sides of the boundary 2-3 ms after the fault is detected is obvious, and the matching error K at the two ends of m and n is obtained through calculation_p0.63 and 0.65 respectively, which are both larger than a locking threshold value K_setWhen the filter is normally put into operation, the two ends of the filter are allowed to protect the outlet, namely 0.2.

As can be seen from FIGS. 9 to 11: within 1-3 ms after the two-end protection detects the fault, i_m-filIs much larger than i_a-filAmplitude of (i)_n-filIs much larger than i_b-filThe amplitude of (d); u. of_m-filHas an amplitude slightly larger than u_a-filIs (about 2 times), u_n-filHas an amplitude slightly larger than u_b-filAmplitude of (about 2 times); p is a radical of_m-filIs much larger than p_a-filAmplitude of p_n-filIs much larger than p_b-filAfter the protection is started, the current and voltage data of a 5ms data window (3.000-3.005 s) are taken, and the identification factors X of the rectifying side and the inverting side are calculated to be 4592, Y is 2.235 multiplied by 10⁴The values are all far larger than the threshold value k equal to 1, so that the protection is determined as the fault in the line area.

As can be seen from FIG. 12, the fault voltage Δ v_m1Is greater than the non-fault voltage av_m2The amplitude of (A) is calculated to obtain that the pole selection function W is 4.39 and is greater than the setting value W_set1Judgment ofAnd the accurate pole selection is realized for the positive pole fault.

There is a document that proposes to identify the internal and external faults of a cell by using the magnitude of a single-ended high frequency quantity in combination with boundary characteristics (e.g. based on line head end high frequency voltage boundary protection), but this method may not be able to protect the full length of the line because: the extra-high voltage direct current line has an attenuation effect on high frequency signals, the longer the line is, the stronger the attenuation effect is, and at the same time, the transition resistance can also reduce the amplitude of high frequency quantity, so that when the tail end of the line has a high resistance ground fault, the high frequency quantity sensed at the protection installation position of the head end (namely, the rectifying end) of the line may be smaller than the high frequency quantity when a fault outside a strong zone at the rectifying side (such as a metallic ground fault outside a smoothing reactor at the rectifying side), as shown in fig. 13. In order to prevent the out-of-area misoperation, the set protection threshold value is larger than the high frequency quantity when the rectifying side is in strong out-of-area fault, so that the scheme cannot protect the whole length of the line.

The invention realizes protection based on the high frequency ratio of two sides of the boundary, does not have the problems and can reliably protect the whole length of the line; moreover, when an out-of-range fault occurs, the fault can be accurately identified only by using single-end electric quantity, and the rapidity is good; the judgment time of single-side protection is less than 10ms when a fault occurs in a region, the inter-station communication time delay is considered, the total protection action time is less than 20ms, and the rapidity is still far superior to that of the conventional current differential protection of a direct-current line.

In order to explore the influence of the fault position and the transition resistance on the protection scheme, the section gives the action condition of protection under the fault condition in different areas, as shown in table 2.

TABLE 2 simulation results under various in-zone fault conditions

Simulation results of the out-of-area faults: when a metallic ground fault occurs outside the inverter-side smoothing reactor (at f3 in fig. 1) (when the dc filter is normally put in use), the protection characteristic amount is as shown in fig. 14 and 15.

As can be seen from fig. 14, after the occurrence of the out-of-range fault on the inversion side, the inversion side rapidly detects the fault signal due to the close distance, and the v side_n-filIs obviously changedLarge, in a period of 3.000-3.005 s, v_n-filIs greater than the threshold value v_setThe side protection is started; the voltage v of the specific frequency band at the rectifying side is influenced by the transmission delay of the line within 5ms after the fault_m-filStill close to 0, v is in the period of 3.005-3.01 s_m-filThe mean absolute value is greater than the threshold value v_setThe side guard is activated. As shown in fig. 15, the waveform difference of the specific frequency band current at two sides of the boundary of 1-2 ms after the fault is detected is obvious, and the matching error K is calculated_p0.41 greater than the latching threshold K_setAllowing the exit to be protected.

As can be seen from fig. 15: within 1-2 ms after the occurrence of the fault outside the inversion side area, i_b-filHas an amplitude slightly larger than i_n-filIs (about 2 times), u_b-filIs much larger than u_n-filAmplitude of p_b-filIs much larger than p_n-filConsistent with theoretical analysis. After the element is started, the current and voltage data of 5ms after the fault occurs are used to obtain the identification factor Y which is 1.8 multiplied by 10^-5And when the value is far smaller than the threshold value k which is equal to 1, the size of X is not needed to be considered, and the fault outside the inversion side area can be quickly judged by only utilizing the single-end fault information.

TABLE 3 simulation results under various out-of-area fault conditions

In order to verify the reliability of the proposed protection scheme in case of an out-of-range fault, the behavior of protection under different out-of-range fault conditions is further given, as shown in table 3. As can be seen from table 3, when a fault occurs outside the area, X or Y is much smaller than the threshold value, the protection reliably identifies the fault outside the area, and at this time, the protection can be implemented only by using single-ended fault information.

The results show that the protection scheme provided by the invention can quickly and reliably identify the faults inside and outside the area, has good resistance to transition resistance, and can accurately select the fault pole when the faults inside the area occur.

The invention provides a protection scheme based on specific frequency band energy ratio values at two sides of a line boundary by theoretical analysis of fault component additional networks in and out of a direct current transmission line area on the basis of impedance frequency characteristics of a direct current filtering link. When the fault occurs in the area, the energy ratios of the specific frequency bands on two sides of the boundary of two ends of the line are both numbers far larger than 1; when the fault occurs outside the area, the specific frequency band energy ratio on two sides of the boundary of the fault end is far less than 1, and the fault inside and outside the area is reliably judged based on the characteristic difference.

The protection scheme can reliably protect the whole length of the line and has good transition resistance; the principle is simple, the realization is easy, the calculated amount is small, and the rapidity is far superior to that of the conventional differential current protection; the high-order energy ratio is adopted in the identification criterion to amplify the difference of the internal and external fault characteristics of the area, so that the threshold value is easy to set, and the reliability of protection under various internal and external fault conditions is effectively ensured.

Finally, it should be pointed out that: the above examples are only for illustrating the technical solutions of the present invention, and are not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A protection method for a high-voltage direct-current transmission line is characterized by comprising the following steps:

step 1: determining criteria for identifying faults inside and outside the area; when the direct current transmission line has an internal fault, the direct current transmission line is judged to have the internal fault by using that the identification factors X, Y at the two ends of the rectifying side and the inverting side are both larger than a criterion threshold value k; when the rectifying side has an external fault, the rectifying end identification factor X is smaller than an identification criterion threshold value k and is directly judged as the external fault; when the inversion side out-of-area fault occurs, the inversion end identification factor Y is smaller than a criterion threshold value k and is directly judged as the out-of-area fault;

in the step 1, measurement points m, n, a, and b are defined, where m and a are respectively line side and valve side measurement points of a line rectification side filtering link, n and b are respectively line side and valve side measurement points of a line inversion side filtering link, and i-order transient energy values of specific frequency bands of each point are respectively:

wherein:

in the formula, p_m-fil、p_n-fil、p_a-fil、p_b-filRespectively sampling values of 2.5-5 kHz frequency band transient power at the measuring points m, n, a and b, wherein the values are equal to corresponding current and voltage fault components i_m-fil、i_n-fil、i_a-fil、i_b-fil、u_m-fil、u_n-fil、u_a-fil、u_b-filThe product of the two; sample number J equals 1, 2, …, J; j is the number of sampling points within 5 ms; i represents the order of the transient energy value, E when i is 4⁽⁴⁾ _m-filRepresenting the 4-order transient energy of m point 2.5-5 kHz frequency band, and the value of the energy is equal to p in the data window_m-filThe sum of the absolute values of the fourth power;

according to the formulas (15) and (16), p is set in the event of an intra-region failure in a high-frequency band_m-fil(j)/p_a-fil(j)>>1，p_n-fil(j)/p_b-fil(j)>>1; in case of an external fault in the rectifying side zone, p_m-fil(j)/p_a-fil(j)<<1; in case of an external fault on the inversion side, p_n-fil(j)/p_b-fil(j)<<1；

Step 2: determining a structural locking criterion;

and step 3: determining a fault starting criterion;

and 4, step 4: determining fault pole selection criteria;

2. The protection method for the hvdc transmission line in accordance with claim 1, wherein: and extracting current and voltage fault components of a specific frequency band by using a Chebyshev filter, wherein the current and voltage fault components are extracted within a time period from the fault to the completion of the regulation of a control system.

3. The protection method for the hvdc transmission line in accordance with claim 1, wherein: the construction locking criterion of the step 2 is based on the waveform matching error K shown in a formula (17)_pTo measure the difference in current waveform:

in the formula, x and y are m, a or n and b respectively representing i_m-filAnd i_a-filOr i_n-fil、i_b-filThe constructed locking criterion is as follows (18):

4. The protection method for the hvdc transmission line in accordance with claim 1, wherein: the fault starting criterion of the step 3 is as shown in a formula (19):

5. The protection method for the hvdc transmission line in accordance with claim 1, wherein: and step 4, determining a pole selection criterion of the formula (21) according to the constructed pole selection function of the formula (20):