A Measurement Abnormality Self-Correction Method for the Digital Twin Protection of Flexible DC Transmission Lines

Abstract

The present invention is a method for self-correcting measurement anomalies for flexible DC line digital twin protection, including the following steps: After performing pole-mode transformation on the instantaneous values of voltage and current at both ends of the flexible DC line, substitute the voltages and currents of the zero mode and line mode into the established state estimation equation. Through state estimation calculation, the sum of normalized squares of residuals can be obtained. If the sum of normalized squares of residuals is less than the residual threshold, let Q(t)=0, then it is regarded as the system operating normally, and step two is executed; if the sum of normalized squares of residuals is greater than the residual threshold, let Q(t)=1, then step three is executed. The present invention can quickly distinguish bad data from fault data, perform corresponding state estimation methods on different data respectively, has the function of bad data identification, thus preventing misoperation of protection caused by bad data and improving the reliability of flexible DC line digital twin protection.

Description

Technical Field

[0001] The present invention belongs to the field of relay protection of power systems, and particularly relates to a method for self-correcting measurement anomalies for digital twin protection of flexible DC lines. Background Art

[0002] As a new generation of revolutionary DC transmission technology, flexible DC transmission technology has gradually been applied in multiple scenarios such as new energy grid connection, power system networking, and island power transmission due to its many advantages such as flexible control, strong controllability, easy networking, small floor area, and no commutation failure, and has good development prospects. However, compared with conventional DC, the flexible DC system has low inertia, small damping, and weak overcurrent tolerance. After a fault occurs in the DC line, the fault current rises sharply, which is likely to cause damage to DC protection and control equipment, and in severe cases, will cause the DC system to shut down. The Zhangbei flexible DC grid demonstration project usually requires the line protection to complete fault discrimination selectively within 3 ms.

[0003] Currently, the protection configuration scheme for flexible DC transmission systems uses traveling wave protection and sudden change protection as the main protection, and pilot differential protection as the backup protection. Traveling wave protection uses the amplitude information of traveling waves to quickly complete fault discrimination and can meet the requirement of rapidity, but its ability to withstand transition resistance is weak. Sudden change protection constructs a protection criterion based on the mutation rate of electrical quantities to perform fault discrimination. The protection action speed is fast, but it also has problems such as low ability to withstand transition resistance and poor anti-interference ability. Pilot differential protection uses the difference in line currents at both ends to perform fault discrimination, with strong selectivity and sensitivity, but it relies on communication and is easily affected by distributed capacitance. Based on the advantages and disadvantages of the above protection schemes, existing scholars have proposed a flexible DC line protection method based on digital twin. This method is based on the frequency variation law of DC lines, establishes a mathematical model corresponding to the actual line in the frequency domain, then performs state estimation on the mathematical model, compares the measured value and the estimated value residual of the electrical quantity on the line, and uses probability and statistics theory to analyze this residual to determine whether the electrical quantity matches the mathematical model, and then discriminates the fault. This protection method has strong selectivity, good ability to withstand transition resistance, and is less affected by lightning interference and noise interference. Incorporating the digital twin concept into DC line protection realizes protection without setting values.

[0004] However, the flexible DC line protection method based on digital twin depends on the accuracy of measurement data. In an actual system, measurement data is easily affected by anomalies such as current transformer anomalies. Abnormal measurement data will affect the result of state estimation, and may even lead to misoperation of the protection. Therefore, this protection method needs to have the ability to quickly detect bad data.

[0005] Existing bad data detection and identification methods for power systems mainly include methods such as neural networks, clustering analysis methods, dynamic inference chains, rough set theory, and Petri net analysis methods. For DC protection control equipment, these methods are too complex in theory and their real-time performance does not meet the quick-acting requirements of DC protection. Therefore, they are not suitable for improving the reliability of the flexible DC line protection method based on digital twins. Summary of the Invention

[0006] Aiming at the defects existing in the flexible DC line protection method based on digital twins, the purpose of the present invention is to provide a measurement anomaly self-correction method for the digital twin protection of flexible DC lines, which can distinguish bad data from fault data, perform corresponding state estimation methods on different data respectively, has the function of bad data identification, thereby preventing misoperation of protection caused by bad data and improving the reliability of protection.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] A measurement anomaly self-correction method for the digital twin protection of flexible DC lines includes the following steps:

[0009] 1) After performing polar mode transformation on the instantaneous values of voltage and current at both ends of the flexible DC line, substitute the voltages and currents of the zero mode and line mode into the established state estimation equation. Through state estimation calculation, the sum of normalized squares of residuals can be obtained. If the sum of normalized squares of residuals is less than the residual threshold, let , then it is considered that the system is operating normally, and step 2) is executed; if the sum of normalized squares of residuals is greater than the residual threshold, let , then step 3) is executed.

[0010] 2) Set the abnormal voltage threshold and the abnormal current threshold , and calculate the difference between the voltage and the rated voltage , and calculate the difference between the current and the rated current . If all voltages and currents satisfy T set at consecutive times, then there is no abnormal data, let , and the state estimation equation remains or returns to the initial state estimation equation (if the state estimation equation was corrected and modified at the previous moment, it needs to return to the initial state estimation equation; if the state estimation equation was not corrected and modified, it remains unchanged); if all voltages and currents do not satisfy T set at consecutive times, let , then keep the state estimation equation unchanged from the state estimation equation of the previous moment, sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step 1);

[0011] 3) Set the abnormal voltage threshold and the abnormal current threshold , and calculate the difference between the voltage and the rated voltage respectively , and calculate the difference between the current and the rated current . Set the abnormal voltage threshold and the abnormal current threshold If the differences between the voltage and current of the same pole at one end of the line and the rated values are simultaneously satisfied , then execute step 4); if the differences between the voltage and current of the same pole at one end of the line and the rated values are not satisfied , then execute step 5);

[0012] 4) If there are 5 or more normalized sums of squared residuals greater than the residual threshold for T set consecutive moments, it is determined that a fault has occurred and the protection operates. If there are less than 5 normalized sums of squared residuals greater than the residual threshold for T set consecutive moments, sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step 1);

[0013] 5) If the differences between the voltage and current of the same pole at one end of the line and the rated values are simultaneously satisfied , then an abnormal voltage appears, let , and execute step 6); if the differences between the voltage and current of the same pole at one end of the line and the rated values are simultaneously satisfied , let , then an abnormal current appears, and execute step 6); if the differences between the voltage and current of the same pole at one end of the line and the rated values are simultaneously satisfied , let , then a disturbance appears, and execute step 7);

[0014] 6) Delete the measured quantities affected by abnormal data in the state estimation, correct the state estimation equation, sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step 1);

[0015] 7) Equate the normalized sum of squared residuals obtained from the state estimation at the current moment to the normalized sum of squared residuals at the previous moment, let , sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step 1);

[0016] Preferably, the difference between the voltage and the rated voltage , and calculate the difference between the current and the rated current The calculation formula is:

[0017] ;

[0018] Wherein is voltage, is rated voltage, is current, is rated current.

[0019] Preferably, the T set setting principle:

[0020] ;

[0021] Preferably, the abnormal voltage threshold :

[0022] ;

[0023] Wherein, is the positive electrode rated voltage, is the minimum abnormal offset coefficient.

[0024] Preferably, the abnormal current threshold :

[0025] ;

[0026] The measurement anomaly self - correction method for the flexible DC line digital twin protection according to the present invention can quickly distinguish bad data from fault data, perform corresponding state estimation methods on different data respectively, has the function of bad data identification, thus preventing misoperation of protection caused by bad data and improving the reliability of the flexible DC line digital twin protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The present invention has the following drawings:

[0028] Figure 1 Equivalent circuit diagram of the flexible DC line modal frequency - dependent model;

[0029] Figure 2 Two - terminal flexible DC line;

[0030] Figure 3 Flow chart of the measurement anomaly self - correction method for the flexible DC line digital twin protection;

[0031] FIG. 4 is a schematic diagram of the comparison result of the influence on protection discrimination before and after self - correction during voltage disturbance;

[0032] FIG. 5 is a schematic diagram of the comparison result of the influence on protection discrimination before and after self - correction during current disturbance;

[0033] FIG. 6 is a schematic diagram of the comparison result of the influence on protection discrimination before and after self - correction during voltage anomaly;

[0034] Figure 7 is a schematic diagram of the comparison results of the influence on protection discrimination before and after self-correction when the current is abnormal;

[0035] Figure 8 is a schematic diagram of the results of the influence on protection discrimination before and after self-correction when data is lost;

[0036] Figure 9 is a schematic diagram of the results of the influence on protection discrimination when an external fault occurs after self-correction when data is lost. Description of the drawings:

[0038] : Positive voltage at the R end of the line;

[0039] : Negative voltage at the R end of the line;

[0040] : Positive voltage at the L end of the line;

[0041] : Negative voltage at the L end of the line;

[0042] : Positive current at the R end of the line;

[0043] : Negative current at the R end of the line;

[0044] : Positive current at the L end of the line;

[0045] : Negative current at the L end of the line;

[0046] : Line impedance;

[0047] : Abnormal zero-mode voltage at the R end of the line;

[0048] : Abnormal line-mode voltage at the R end of the line;

[0049] : Voltage at the R end in the mode domain;

[0050] : Reverse traveling wave of voltage at the L end of the line;

[0051] : Reverse traveling wave of voltage at the R end of the line;

[0052] : Abnormal voltage at the R end in the mode domain;

[0053] : Abnormal state quantity;

[0054] : The Jacobian matrix of the measurement function;

[0055] : The weight matrix, and the coefficients on its diagonal reflect the noise level of each measurement quantity;

[0056] : The historical value matrix composed of state variables;

[0057] : The estimated value of the measurement quantity indicating an anomaly;

[0058] : The normalized sum of squared residuals indicating an anomaly;

[0059] : The relationship matrix composed of known parameters of the line;

[0060] It means that the normalized sum of squared residuals is greater than or equal to the residual threshold; The normalized sum of squared residuals is less than the residual threshold;

[0061] It means the residual threshold;

[0062] It means that the relay issues a trip command; It means that the relay does not issue a trip command;

[0063] : The protection criterion time window;

[0064] : The protection threshold;

[0065] : The positive pole voltage;

[0066] : The rated voltage of the positive pole is 250 kV;

[0067] : The negative pole voltage;

[0068] : The maximum fluctuation range of the capacitor voltage of the converter sub-module;

[0069] : The difference between the absolute values of the positive pole voltage and the negative pole voltage;

[0070] : The offset value of the measured voltage from the rated voltage;

[0071] : The threshold of abnormal voltage data;

[0072] : The rated current of the line is 1.25 kA;

[0073] : The minimum current value for a fault within the zone;

[0074] : The absolute value of the deviation of the measured current from the rated current;

[0075] : The threshold value of the abnormal current data.

[0076] Indicates that the state estimation equation is maintained or restored to the initial state estimation equation; Indicates that the state estimation equation remains unchanged from the previous moment's state estimation equation;

[0077] Indicates correcting the abnormal voltage; Indicates correcting the abnormal current; , indicating correcting the disturbance data. Specific implementation manner

[0078] The following further elaborates on the present invention in conjunction with Figure 1 -9.

[0079] 1. Bad data affects the discrimination of the digital twin protection of the flexible DC line

[0080] In the digital twin protection of the flexible DC line, the measured quantities are divided into the true measured quantities after phase-mode transformation of the actually measured electrical quantities and the virtual measured quantities that satisfy physical laws. Bad data only affects the true measured quantities. The virtual measured quantities represent physical laws and there will be no measurement anomaly problems. When the actually measured electrical quantity is bad data, it will first affect the electrical quantities such as the line mode and zero mode after phase-mode transformation. For example Figure 2 the positive voltage at the R end of the line becomes bad data , the voltage measurement quantity after phase-mode transformation is:

[0081] ;

[0082] In the formula: are the abnormal zero-mode and line-mode voltages at the R end of the line respectively; is the negative voltage at the R end of the line.

[0083] It can be seen that when the voltage or current data of any pole is bad data, it will cause the voltage or current of the line mode and zero mode to be abnormal.

[0084] Secondly, the abnormal true measured quantity will affect the estimated value of the virtual measured quantity. For example Figure 1The reverse traveling wave of the voltage at the L end of the line is equal to the forward traveling wave of the voltage at the R end. When the voltage at the R end of the line in the modal domain becomes abnormal data , according to the convolution theorem, the physical relationship can be obtained:

[0085] ;

[0086] In the formula: the estimated value of the virtual measurement quantity is , which is no longer equal to 0, indicating that the state estimation equation does not hold and no longer follows the physical law; The reverse traveling wave of the voltage at the L end of the line; The reverse traveling wave of the terminal voltage; are all constants and can be obtained by the convolution theorem and calculation.

[0087] At this time, the measurement vector constructed by the abnormal true measurement quantity and the virtual measurement quantity is also abnormal, and the state quantity obtained by its state estimation is also abnormal. It can be obtained that:

[0088] ;

[0089] In the formula: is the abnormal state quantity; is the Jacobian matrix of the measurement function; is the weight matrix, and the coefficients on its diagonal reflect the noise level of each measurement quantity; is the historical value matrix composed of state quantities;

[0090] It can be analyzed from formula (3) that historical data will affect the estimated value of measurement data in subsequent state estimation.

[0091] Finally, the estimated value of the measurement quantity and the normalized sum of squared residuals also become abnormal data, as shown in the following formula:

[0092] ;

[0093] ;

[0094] In the formula: represents the estimated value of the abnormal measurement quantity; represents the normalized sum of squared residuals of the abnormal residuals; is the relationship matrix composed of known parameters of the line.

[0095] At this time, the abnormal normalized sum of squared residuals may be greater than the residual threshold, and the relay issues a tripping action, as shown in the following formula:

[0096] ;

[0097] In the formula: Indicates that the residual normalized sum of squares is greater than or equal to the residual threshold; Indicates the residual threshold; Indicates that the relay issues a trip command; Is the protection criterion time window; Is the protection threshold;

[0098] It can be analyzed from Equation (6) that bad data may cause misoperation of protection based on state estimation, and state estimation needs to calculate the measurement data at each sampling moment. Therefore, it is necessary to identify bad data and fault data in real time.

[0099] 2. Identify and correct bad data and fault data

[0100] When the abnormal residual normalized sum of squares is greater than the residual threshold, it means that there is a fault or bad data at this time. If there is fault data, the electrical quantities of the fault pole will change greatly at the same time; if there is bad data, it usually shows that individual data is abnormal. Therefore, the amplitude and type of the change of the measurement data can be used as features to distinguish bad data and fault data.

[0101] According to the amplitude of the deviation of the bad data from the rated value, the bad data can be divided into disturbance data with less impact on the system and abnormal data with greater impact on the system. The following is to distinguish disturbance data and abnormal data.

[0102] Assume that the positive and negative line parameters are symmetric and the line impedance is also symmetric, then the electrical quantities at both ends should also be symmetric. It is possible to judge whether there is bad data by judging whether the electrical quantities of the two poles are symmetric. Taking voltage as an example, when the system is operating normally and the measuring element is normal, the positive and negative pole voltages of the line are:

[0103] ;

[0104] In the formula: Is the positive pole voltage; Is the rated positive pole voltage of 250 kV; Is the negative pole voltage; Are both voltage offset coefficients; Is the maximum fluctuation range of the capacitor voltage of the converter sub-module.

[0105] When the measuring element is working normally, the measured values of the positive pole voltage and the negative pole voltage are symmetric, and the absolute value difference between the measured values of the two pole voltages can be expressed by Equation (8):

[0106] ;

[0107] In the formula: Is the difference between the absolute values of the positive pole voltage and the negative pole voltage, which can represent the symmetric offset value of the positive and negative pole voltages.

[0108] Combining equations (7) and (8) gives the normal fluctuation range of:

[0109] ;

[0110] Assume that the measured value of the positive electrode voltage is abnormal, and the abnormal offset coefficient is , then the positive electrode voltage is , then there is:

[0111] ;

[0112] When the measured value of the positive electrode voltage is abnormal, the positive and negative electrode voltage values are no longer symmetric. Combining equations (9) and (10), it can be obtained that:

[0113] ;

[0114] The minimum abnormal offset coefficient can be obtained . When , at this time, the symmetric offset value of the positive and negative electrode voltages has exceeded the normal range, and the positive and negative electrode voltages are asymmetric.

[0115] Therefore, when the voltage measurement data is abnormal, its voltage offset value is as shown in equation (12).

[0116] ;

[0117] In the formula: is the measured voltage; represents the offset value between the measured voltage and the rated voltage.

[0118] And when , at this time, is regarded as disturbance data. Set the threshold value of the abnormal voltage data, then the abnormal voltage criterion is as shown in equation (13).

[0119] ;

[0120] For current data, since there is no current offset standard, the threshold value of the fault location auxiliary criterion is used here to judge whether the current is abnormal, as follows:

[0121] ;

[0122] In the formula: is the measured current; is the rated current of the line, 1.25 kA; the minimum current value for internal faults; is the absolute value of the measured current offset from the rated current.

[0123] And when , at this time, is regarded as disturbance data. Set the threshold of abnormal current data , then the abnormal current criterion is shown in Equation (15):

[0124] ;

[0125] Then, using the abnormal data criterion, the fault data and bad data can be distinguished. If the voltage and current of the same pole at one end of the line both satisfy the abnormal data criterion for the fault data, the fault startup criterion can be set as shown in Equation (16):

[0126] ;

[0127] The bad data is further divided into disturbance data and abnormal data. The abnormal data is further divided into abnormal voltage data and abnormal current data. The abnormal voltage data is shown in Equation (17):

[0128] ;

[0129] The abnormal current data is shown in Equation (18):

[0130] ;

[0131] The disturbance data is shown in Equation (19):

[0132] ;

[0133] For the fault data that satisfies the fault startup criterion, the sum of normalized squares of residuals needs to be submitted to the protection criterion to determine whether the line has a real fault.

[0134] The bad data is further divided into disturbance data and abnormal data. The abnormal data is further divided into abnormal voltage data and abnormal current data. For the disturbance data, it can be regarded as being slightly affected during the data measurement or communication process, and has little impact on the historical values in the state estimation. The sum of normalized squares of residuals estimated at the current moment can be made equal to the value at the previous moment. For the abnormal data, its amplitude changes greatly, and it may be due to abnormal current transformers, etc. At this time, the measured quantities affected by the abnormal data can be deleted, and the measured data at the opposite end of the line can be added as the measured quantities, so as to correct the state estimation equation and historical values, facilitating the next state estimation.

[0135] 3. The flowchart of a measurement anomaly self-correction method for DC flexible transmission line digital twin protection provided by the present invention includes the following steps:

[0136] 1) After performing the pole-mode transformation on the instantaneous values of the voltage and current at both ends of the flexible DC line, substitute the voltages and currents of the zero mode and line mode into the established state estimation equation. Through state estimation calculation, the sum of the normalized squares of the residuals can be obtained. If the sum of the normalized squares of the residuals is less than the residual threshold, let , then it is regarded that the system is operating normally, and step 2) is executed; if the sum of the normalized squares of the residuals is greater than the residual threshold, let , then step 3) is executed.

[0137] 2) Set the abnormal voltage threshold and the abnormal current threshold . Calculate the difference between the voltage and the rated voltage respectively, and calculate the difference between the current and the rated current . If all voltages and currents satisfy T set continuously for time, then there is no abnormal data, let , and the state estimation equation remains or is restored to the initial state estimation equation (if the state estimation equation was corrected and modified at the previous moment, it needs to be restored to the initial state estimation equation; if the state estimation equation was not corrected and modified, it remains unchanged); if all voltages and currents do not satisfy T set continuously for , let , then keep the state estimation equation unchanged from the state estimation equation of the previous moment, sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step 1);

[0138] 3) Set the abnormal voltage threshold and the abnormal current threshold . Calculate the difference between the voltage and the rated voltage respectively, and calculate the difference between the current and the rated current . Set the abnormal voltage threshold and the abnormal current threshold If it is detected that the differences between the voltage and current of the same pole at one end of the line and the rated values simultaneously satisfy , then step 4) is executed; if the differences between the voltage and current of the same pole at one end of the line and the rated values do not satisfy , then step 5) is executed;

[0139] 4) If there are 5 or more sums of the normalized squares of the residuals greater than the residual threshold continuously for T set time, then it is judged that a fault has occurred and the protection action is taken. If continuously for T setThere are less than 5 sum of normalized squares of residuals greater than the residual threshold at a moment. Sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step 1);

[0140] 5) If the differences between the voltage and current of the same pole at one end of the line and the rated value simultaneously satisfy , then an abnormal voltage appears. Let , and execute step 6); If the differences between the voltage and current of the same pole at one end of the line and the rated value simultaneously satisfy , let , then an abnormal current appears, and execute step 6); If the differences between the voltage and current of the same pole at one end of the line and the rated value simultaneously satisfy , let , then a disturbance appears, and execute step 7);

[0141] 6) Delete the measured quantities affected by abnormal data in the state estimation, correct the state estimation equation, sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step 1);

[0142] 7) Equalize the sum of normalized squares of residuals obtained from the state estimation at the current moment to the sum of normalized squares of residuals at the previous moment. Let , sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step 1);

[0143] Figure 4 is a schematic diagram of the comparison results of the influence on protection discrimination before and after self-correction during voltage disturbance provided by an embodiment of the present invention. Figure 4(a) shows that a ground fault occurs when the positive pole in the zone is grounded through a 300 transition resistance at 5 ms. Since the protection threshold is that the sum of normalized squares of residuals at 5 sampling moments within 1 ms is greater than the threshold, disturbance voltage data of the positive pole voltage at side L are respectively set at ten consecutive sampling points within 0 ms to 1 ms. The disturbance voltage deviates from the rated value by 8%. The influence of the disturbance voltage on protection discrimination is shown in Figure 4(b); the influence of the disturbance voltage on protection discrimination after measurement self-correction is shown in Figure 4(c). During the fault transient, compared with the current, the voltage exceeds the disturbance threshold faster, , and the state estimation equation is corrected according to the abnormal voltage. After the fault current exceeds the disturbance upper limit, the residuals obtained from the state estimation will be used for protection discrimination. By comparing and analyzing Figure 4(b) and Figure 4(c), after adaptive error correction, it can withstand the interference of disturbance voltage data and improve the reliability of protection discrimination.

[0144] Figure 5 is a schematic diagram of the comparison results of the influence on protection discrimination before and after self-correction during current disturbance provided by an embodiment of the present invention. Figure 5(a) shows that a ground fault occurs when the positive pole in the zone is grounded through a 300 For the ground fault with transitional resistance, since the protection threshold is that the sum of the normalized squares of the residuals at 5 sampling moments within 1 ms is greater than the threshold, the disturbing data of the positive-pole current on the L side are respectively set at ten consecutive sampling points within 0 ms to 1 ms. The disturbing current deviates from the rated value by 8% in all cases. The influence of the disturbing current on the protection discrimination is shown in Fig. 5(b); after measurement self-error correction, the influence of the disturbing current on the protection discrimination is shown in Fig. 5(c). During the fault transient, compared with the current, the voltage exceeds the disturbance threshold faster. , according to the abnormal voltage correction state estimation equation, after the fault current exceeds the disturbance upper limit, the residuals obtained from the state estimation will be used for protection discrimination. By comparing and analyzing Fig. 5(b) and Fig. 5(c), after adaptive error correction, it can withstand the interference of the disturbing current data, improving the reliability of protection discrimination.

[0145] Fig. 6 is a schematic diagram of the comparison results of the influence on protection discrimination before and after self-error correction when the voltage is abnormal provided by the embodiment of the present invention. Fig. 6(a) shows that a ground fault with the positive pole in the zone through a 300 For the ground fault with transitional resistance, since the protection threshold is that the sum of the normalized squares of the residuals at 5 sampling moments within 1 ms is greater than the threshold, the abnormal voltages of the positive-pole voltage on the L side are respectively set at ten consecutive sampling points within 0 ms to 1 ms. The abnormal voltages deviate from the rated value by 20% in all cases. The influence of the abnormal voltages on the protection discrimination is shown in Fig. 6(b); after measurement self-error correction, the influence of the abnormal voltages on the protection discrimination is shown in Fig. 6(c). During the fault transient, compared with the current, the voltage exceeds the disturbance threshold faster. , according to the abnormal voltage correction state estimation equation, after the fault current exceeds the disturbance upper limit, the residuals obtained from the state estimation will be used for protection discrimination. By comparing and analyzing Fig. 6(b) and Fig. 6(c), after adaptive error correction, it can withstand the interference of the abnormal voltage data, improving the reliability of protection discrimination.

[0146] Fig. 7 is a schematic diagram of the comparison results of the influence on protection discrimination before and after self-error correction when the current is abnormal provided by the embodiment of the present invention. Fig. 7(a) shows that a ground fault with the positive pole in the zone through a 300 For the ground fault with transitional resistance, since the protection threshold is that the sum of the normalized squares of the residuals at 5 sampling moments within 1 ms is greater than the threshold, the abnormal data of the positive-pole current on the L side are respectively set at ten consecutive sampling points within 0 ms to 1 ms. The abnormal currents deviate from the rated value by 20% in all cases. The influence of the abnormal currents on the protection discrimination is shown in Fig. 7(b); after measurement self-error correction, the influence of the abnormal currents on the protection discrimination is shown in Fig. 7(c). During the fault transient, compared with the current, the voltage exceeds the disturbance threshold faster. , according to the abnormal voltage correction state estimation equation, after the fault current exceeds the disturbance upper limit, the residuals obtained from the state estimation will be subjected to protection discrimination. By comparing and analyzing Fig. 7(b) and Fig. 7(c), after adaptive error correction, it can withstand the interference of abnormal current data and improve the reliability of protection discrimination.

[0147] Fig. 8 is a schematic diagram of the influence of self-error correction before and after data loss on protection discrimination provided by an embodiment of the present invention. When abnormal conditions such as disconnection or fault occur in the mutual inductor, data loss is very likely to occur at this time. At this time, the measured data is 0 and deviates from the rated value of 100% for a long time, which belongs to abnormal data and will cause misoperation of the protection. In Fig. 8(a), voltage data loss is set at 0 ms in the simulation, and in Fig. 8(b), current data loss is set at 0 ms in the simulation. Both Fig. 8(a) and Fig. 8(b) set a ground fault of the positive pole in the zone through a 300 transition resistance at 5 ms. As can be seen from Fig. 8(a) and Fig. 8(b), before the fault, after the measured data loss is subjected to adaptive error correction, the state estimation equation is continuously corrected by using the redundancy of the measured quantities, and the fault can be accurately identified, improving the reliability of protection discrimination.

[0148] Fig. 9 is a schematic diagram of the influence of an external fault on protection discrimination after self-error correction when data is lost provided by an embodiment of the present invention. When self-error correction is performed after data loss, the state equation has been corrected at this time, and the influence of an external fault on protection discrimination needs to be verified. In Fig. 9(a), voltage data loss is set at 0 ms in the simulation, and in Fig. 9(b), current data loss is set at 0 ms in the simulation. Both Fig. 9(a) and Fig. 9(b) set an external bipolar short-circuit fault at 5 ms. As can be seen from Fig. 9(a) and Fig. 9(b), after an external fault occurs, the sum of the normalized squares of the residuals fluctuates slightly but does not exceed the threshold, and the protection discrimination is not affected, indicating that the state estimation equation still follows the physical laws on the line after correction, can correctly identify internal and external faults, and the measurement anomaly self-error correction method for the digital twin protection of flexible DC transmission lines improves the reliability of protection discrimination.

[0149] The content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

Claims

1. A measurement anomaly self-correction method for flexible DC line digital twin protection, characterized in that Including the following steps: Step 1: After performing polar mode transformation on the measured voltage and instantaneous value of the measured current at both ends of the flexible DC line, substitute the voltages and currents of the zero mode and line mode into the established state estimation equation. Through state estimation calculation, obtain the normalized sum of squares of residuals. If the normalized sum of squares of residuals is less than the residual threshold, let , then it is regarded that the system is operating normally, and proceed to Step 2; if the normalized sum of squares of residuals is greater than the residual threshold, let , then proceed to Step 3; Step 2: Set the abnormal voltage threshold and the abnormal current threshold , and calculate the difference between the measured voltage and the rated voltage , and calculate the difference between the measured current and the rated current ; If continuous T set At all times, all voltages and currents satisfy , then there is no abnormal data. Let , the state estimation equation remains or is restored to the initial state estimation equation. If the state estimation equation was corrected and modified at the previous moment, it needs to be restored to the initial state estimation equation. If the state estimation equation was not corrected and modified, it remains unchanged; If continuous T set At all times, all voltages and currents do not satisfy , let , then keep the state estimation equation unchanged from the previous moment, sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat step one; Step 3: Set the abnormal voltage threshold and the abnormal current threshold , calculate the difference between the measured voltage and the rated voltage , calculate the difference between the measured current and the rated current ; Set the abnormal voltage threshold and the abnormal current threshold If the differences between the voltage and current of the same pole at one end of the line and their rated values are both satisfied , then execute Step 4; if the differences between the voltage and current of the same pole at one end of the line and their rated values are not satisfied , then execute Step 5; Step 4: If there are 5 or more sum of normalized squares of residuals greater than the residual threshold for T set successive moments, a fault is determined and the protection operation is performed; if there are less than 5 sum of normalized squares of residuals greater than the residual threshold for T set successive moments, the voltage and current signals at both ends of the flexible DC line at the next moment are sampled, and Step 1 is repeated; Step 5: If the differences between the voltage and current of the same pole at one end of the line and their rated values simultaneously satisfy , an abnormal voltage occurs. Let , and execute Step 6; if the differences between the voltage and current of the same pole at one end of the line and their rated values simultaneously satisfy , let , an abnormal current occurs, and execute Step 6; if the differences between the voltage and current of the same pole at one end of the line and their rated values simultaneously satisfy , let , a disturbance occurs, and execute Step 7; Step 6: Delete the measurement quantities affected by abnormal data in the state estimation, correct the state estimation equation, sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat Step 1; Step 7: Equalize the sum of normalized squares of the residuals obtained from the current moment state estimation to the sum of normalized squares of the residuals at the previous moment, and let , sample the voltage and current signals at both ends of the flexible DC line at the next moment, and repeat Step 1.

2. The measurement anomaly self-correction method for the digital twin protection of the flexible DC line according to claim 1, characterized in that Calculate the difference between the measured voltage and the rated voltage , and calculate the difference between the measured current and the rated current The calculation formula is as follows: ; wherein is the measured voltage, is the rated voltage, is the measured current, is the rated current.

3. The self - error - correction method for measurement anomalies in a flexible DC line digital twin protection as claimed in claim 1, wherein, The T set setting principle: 。 4. The self - error - correction method for measurement anomalies in a flexible DC line digital twin protection as claimed in claim 1, wherein, The abnormal voltage threshold : ; Among them, is the rated voltage of the positive electrode, is the minimum abnormal offset coefficient.

5. The measurement anomaly self-correction method for flexible DC line digital twin protection according to claim 1, characterized in that The abnormal current threshold : ; Among them, is the rated current of the line, is the minimum current value for internal faults.

Images