A direct current distribution network grounding fault protection method based on positive and negative electrode current correlation

Abstract

The application provides a direct current distribution network grounding fault protection method based on positive and negative electrode current correlation, a VSC based on symmetric single pole wiring, a neutral point grounding branch current used as a protection starting criterion to quickly detect a fault; and based on the characteristic that the symmetry of a system fault loop is destroyed, a Pearson correlation coefficient used as an analysis and calculation tool, a change trend characteristic of positive and negative electrode fault currents used for fault positioning and fault pole selection, a bipolar current principle only using a communication channel to transmit a logic signal, and a low requirement for the synchronization and channel capacity of a communication system, while ensuring the reliability, transition resistance tolerance and noise capacity of the protection, and saving the cost of communication equipment. The method has low requirement for communication equipment, needs less communication resources, and does not require strict two-end data synchronization.

Description

Technical Field

[0001] This invention relates to the field of power grid distribution and automation technology, specifically a method for protecting DC distribution network grounding faults based on the correlation between positive and negative currents. Background Technology

[0002] Compared to AC power grids, DC power grids offer advantages such as high efficiency in integrating distributed power sources, ease of grid connection, large power supply capacity, and strong anti-interference performance, making them highly valuable for research and application. In recent years, DC power grid protection technology has been widely applied in medium- and low-voltage distribution network projects, effectively adapting to the new scenario of widespread distributed power source integration and profoundly changing the future development pattern of the power grid. However, current low-voltage DC distribution network protection technology still faces many technical challenges in fault detection, fault location, and fault recovery. Therefore, rationally utilizing and resolving the profound contradictions among the four characteristics of protection has become a key issue in ensuring the safe and stable operation of the power grid.

[0003] A search revealed a related patent document, Chinese Patent Publication No. CN115632445A, which provides a protection method for AC / DC hybrid distribution networks based on 5G communication. For DC grounding faults in AC / DC hybrid distribution networks, the method detects the occurrence of the grounding fault by utilizing the sum of the total currents at each protection installation point in the DC subnet. Then, based on 5G communication, it obtains the sum of the total currents of each protection element on the downstream adjacent lines and determines the fault section based on the amplitude relationship. Furthermore, a capacitive polarity selection factor is introduced to reliably select the fault polarity based on the magnitude of the DC bipolar current of the voltage source converter and the AC side grounding current. The polarity selection results are transmitted using the 5G network, thereby achieving selective tripping of protection and accurate fault isolation. This invention utilizes current amplitude information in AC / DC hybrid distribution networks and 5G high-speed communication to achieve effective identification and location of DC grounding faults. It can adapt to different DC subnet power supply topologies, helping to improve the operational economy and power supply reliability of AC / DC distribution networks.

[0004] However, the power grid environment in which the above-mentioned patents are applied differs from that in this case, and the fault monitoring and location methods used are different from the technical solutions in this case.

[0005] Currently, there are records of using two-terminal current correlation analysis to analyze fault currents in high-voltage direct current (HVDC) transmission lines. Peng Tao et al. proposed a new method for protection based on two-terminal current correlation (Protection of HVDC Transmission Lines Based on Current Correlation Analysis, Journal of Electric Power System and Automation, 2018(030)004). This method compares the correlations based on the characteristics that the current changes in opposite directions at both ends during faults within the fault zone, while the current changes in the same direction at both ends during faults outside the fault zone, thus achieving rapid and accurate identification of faults within and outside the fault zone. This method does not require strict time synchronization of the protection devices at both ends of the line, has low sampling frequency requirements, can be implemented under the current hardware conditions of HVDC control and protection systems, has high reliability and strong practicality, and can meet the protection performance requirements of HVDC lines. PSCAD and MATLAB simulation results verify its feasibility and effectiveness.

[0006] The protection principles of low-voltage DC distribution networks can be divided into two types based on communication methods: single-ended quantity protection principles and protection principles based on two-ended communication. Single-ended quantity protection principles utilize only electrical quantity information from one side to construct protection, achieving fault location without communication. However, they are slightly lacking in selectivity, and some methods heavily rely on line boundary reactance. Protection principles based on two-ended communication directly utilize electrical quantity information from both sides of the component, constructing protection based on longitudinal protection methods. Their implementation is simpler, offering absolute selectivity and strong tolerance to transition resistance, but they require higher synchronization in communication.

[0007] Currently, protection principles based on two-way communication place high demands on communication channels and equipment, and require strict synchronization of electrical quantities on both sides of the line, leading to increased communication costs and failing to meet the economic principles of low-voltage DC systems. Therefore, a novel protection method is needed while ensuring protection reliability. Summary of the Invention

[0008] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention provides a method for protecting DC distribution network grounding faults based on the correlation between positive and negative pole currents.

[0009] The specific technical solution is as follows:

[0010] A method for protecting DC distribution network grounding faults based on the correlation between positive and negative pole currents, comprising the following steps:

[0011] (1) Fault detection, the method is as follows:

[0012] capacitor neutral point grounding branch current I gA sudden change occurs after a fault, which can quickly and accurately reflect the fault and is used as a protection start criterion. After the fault is detected, the IGBT in the system is locked to realize the self-protection of VSC; the IGBT between the three-phase bridge arm of VSC and the DC side output capacitor is quickly turned off, cutting off the connection between the AC and DC sides.

[0013] (2) Fault location, the steps are as follows:

[0014] ① Sample the positive and negative currents at each protection installation point in the system and calculate the Pearson correlation coefficient for each side of each line;

[0015] ②The S side is judged according to the protection criteria. If the S side is judged to be a positive direction fault, an allow signal is sent to the L side and the response is waited for. If the line segment is the end line of the system, it can be directly judged as a fault in the line area based on the judgment result of the S side.

[0016] ③ After receiving the protection signal, if the L side also meets the positive direction fault criteria, it can be determined that the fault is within the line area and an action signal is sent to the S side at the same time; if the L side does not meet the positive direction fault criteria, it is determined that the fault is outside the line area and the protection on this side is blocked.

[0017] ④ After receiving the confirmation signal from the L side, the S side initiates fault polarity selection.

[0018] (3) Fault polarity selection, the method of which is:

[0019] Based on the characteristics of the fault current change on the S side of the line, after determining that the fault occurs in the positive direction, the positive and negative currents on the S side are used as the basis for pole selection. When a positive pole ground fault occurs, the positive current rises rapidly, while the negative current decreases slightly. The amplitude of the positive current will be much higher than that of the negative current. When a negative pole ground fault occurs, the opposite will happen. The corresponding circuit breakers on both sides of the line S and L trip according to the fault pole selection criteria, and the faulty line is disconnected.

[0020] Furthermore, the fault detection described in step (1) involves the capacitor neutral point grounding branch current I. g The activation criterion when a change occurs is shown in the following formula:

[0021] |ΔI g |>ΔI g.set

[0022] Where, |ΔI g | represents the absolute value of the sudden change in current in the neutral-point grounding branch of the capacitor, ΔI g.set The threshold for activation should take into account the actual engineering situation and leave sufficient margin for protection.

[0023] Furthermore, the specific method for calculating the Pearson correlation coefficient on each side of each line in step (2) is as follows: The Pearson correlation coefficient is used to reflect the correlation between the positive and negative current change trends on one side of the line. The Pearson correlation coefficient is defined as follows:

[0024]

[0025] Among them, i P ={x1,x2,…,x n} represents the positive current sampling value, i N ={y1,y2,…,y n} represents the negative electrode current sampling value. is the mean of the sampled data, and m is the number of the two sides (S and L sides) of the line.

[0026] Furthermore, in step (3), the fault selection uses the ratio of the two pole currents as the fault selection criterion, and the current ratio K is defined as follows:

[0027]

[0028] After a positive fault occurs on side S, the current ratio K will be much greater than 1 or much less than 1. Considering the margin required in the system and measurement errors, the fault pole selection criterion is set as follows:

[0029]

[0030] Moreover, the Pearson correlation coefficient P m The value range is [-1, 1], P m When P = 1, there is a perfect positive correlation; m When P = -1, there is a perfect negative correlation; m =0, indicating no correlation.

[0031] Furthermore, for S-side protection, considering the required margin of protection, the protection operation conditions can be set as follows:

[0032] P S <0.7;

[0033] For L-side protection, a blocking element needs to be added, configured to only release the protection after receiving a permission signal from S-side protection. Therefore, the operating conditions for L-side protection can be set as follows:

[0034]

[0035] Furthermore, the current ratio K is taken as the number of sampling points N=10 for each comparison, and the fault pole can be determined by using a sliding time window to calculate three times in a row to satisfy the above criteria.

[0036] Moreover, for the criteria, the calculation window of the Pearson correlation coefficient algorithm is 20 sampling points. If the criteria are met three times consecutively within the sliding window, it can be determined that there is a positive fault in the protection direction on that side, and the protection will be activated.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] 1. This invention is based on a symmetrical single-pole VSC, utilizing the neutral-point grounded branch current as the protection activation criterion for rapid fault detection. Furthermore, based on the characteristic of disrupted system fault circuit symmetry, it employs the Pearson correlation coefficient as an analysis and calculation tool, utilizing the changing trends of the positive and negative fault currents to locate and select fault poles. The bipolar current principle involved only utilizes the communication channel to transmit logic signals, thus requiring low synchronization and channel capacity from the communication system. While ensuring protection reliability and tolerance to transition resistance and noise, it saves on communication equipment costs. This method has low requirements for communication equipment, requires few communication resources, and does not require strict data synchronization between the two ends.

[0039] 2. The DC distribution network grounding fault protection method based on the correlation between positive and negative pole currents provided by this invention has absolute selectivity, can accurately and reliably identify the fault section for grounding faults, has strong resistance to transition resistance, and strong anti-noise capability.

[0040] 3. The DC distribution network grounding fault protection method based on the correlation between positive and negative currents provided by this invention constructs protection according to the positive and negative symmetrical characteristics of the system after the fault, and has strong adaptability to distributed power supply access and system load fluctuations.

[0041] 4. The DC distribution network grounding fault protection method based on the correlation between positive and negative currents provided by this invention samples the positive and negative currents at each protection installation point in the system. Compared with the protection principle based on differential state quantities, it can make more effective use of the sampled information and has higher reliability. Attached Figure Description

[0042] Figure 1 This is a topology diagram of a low-voltage DC system with distributed power supply access.

[0043] Figure 2 An equivalent circuit model for a positive ground fault in a low-voltage DC system.

[0044] Figure 3 This is the fault detection scheme and the judgment logic for IGBT lockout.

[0045] Figure 4 This is the equivalent circuit for the DC-side capacitor discharge stage of VSC.

[0046] Figure 5To protect the action logic flowchart. Detailed Implementation

[0047] The technical solution of the present invention will be further illustrated below with reference to specific embodiments.

[0048] This invention provides a DC distribution network grounding fault protection method based on the correlation between positive and negative currents. The protection principle is constructed by utilizing the abrupt change in the neutral point grounding branch current of the VSC DC side outlet capacitor and the correlation characteristics of the positive and negative currents of the line. The process mainly consists of the following three stages:

[0049] 1) Fault detection

[0050] Due to the grounding branch current I of the VSC DC side outlet capacitor neutral point... g A sudden change occurs after a fault, which can quickly and accurately reflect the fault; therefore, it is used as the protection activation criterion. Considering measurement errors, the activation criterion is set as follows:

[0051] |ΔI g |>ΔI g.set

[0052] Where, |ΔI g | represents the absolute value of the sudden change in current in the neutral-point grounding branch of the capacitor, ΔI g.set The threshold for activation should take into account the actual engineering situation and leave sufficient margin for protection.

[0053] Upon detection of a fault, the IGBTs in the system are locked out, enabling self-protection of the VSC; the IGBTs between the three-phase bridge arms of the VSC and the DC-side output capacitor are quickly turned off, severing the connection between the AC and DC sides.

[0054] 2) Fault location

[0055] Assume a fault occurs at the positive terminal of a line, and designate the side of the line closer to the neutral point of the VSC as the S side and the side farther from the VSC neutral point as the L side. After a DC-side fault occurs, the positive current in the fault circuit (i.e., between the VSC DC-side capacitor outlet and the fault point) rises rapidly, reaches its peak, and then slowly decreases; the negative current first decreases rapidly, then gradually rises, and finally decreases slowly. Therefore, the trends of the positive and negative currents on the S side of the faulty line are different, while the trends of the positive and negative currents on the L side are the same. Based on these fault characteristics, a ground fault protection principle based on the correlation between positive and negative currents can be proposed.

[0056] This invention utilizes the Pearson correlation coefficient to reflect the correlation between the changing trends of the positive and negative currents on one side of a circuit. The Pearson correlation coefficient is defined as follows:

[0057]

[0058] Among them, i P ={x1,x2,…,x n} represents the positive current sampling value, i N ={y1,y2,…,y n} represents the negative electrode current sampling value. is the mean of the sampled data, and m is the number of the two sides (S and L sides) of the line.

[0059] Pearson correlation coefficient P m The value range is [-1, 1]. P m When P = 1, there is a perfect positive correlation; m When P = -1, there is a perfect negative correlation; m =0, indicating no correlation.

[0060] For S-side protection, considering the required margin of protection, the protection operation conditions can be set as follows:

[0061] P S <0.7

[0062] For L-side protection, a blocking element needs to be added, configured to only release the protection upon receiving a permission signal from S-side protection. Therefore, the operating conditions for L-side protection can be set as follows:

[0063]

[0064] For the above criteria, the calculation window of the Pearson correlation coefficient algorithm is 20 sampling points. If the criteria are met three times consecutively under the sliding window length, it can be determined that there is a positive fault in the protection direction on that side, and the protection will be activated.

[0065] The protection action logic during the fault identification phase is as follows:

[0066] (1) Sample the positive and negative currents at each protection installation point in the system and calculate the Pearson correlation coefficient for each side of each line;

[0067] (2) The S side is judged according to the protection criteria. If the S side is judged to be a positive direction fault, an allow signal is sent to the L side and the response is waited for. If the line segment is the end line of the system, it can be directly judged as a fault in the line area based on the judgment result of the S side.

[0068] (3) After receiving the protection signal, if the L side also meets the positive direction fault criteria, it can be judged as a fault within the line area and an action signal is sent to the S side at the same time; if the L side does not meet the positive direction fault criteria, it is judged as a fault outside the line area and the protection on this side is blocked.

[0069] (4) After receiving the confirmation signal from the L side, the S side initiates fault selection.

[0070] 3) Fault polarity selection

[0071] Based on the characteristics of the fault current change on the S side of the line, after determining that the fault occurs in the positive direction, the positive and negative currents on the S side can be used as the basis for pole selection. After a positive ground fault occurs, the positive current rises rapidly, while the negative current decreases slightly, and the amplitude of the positive current will be much higher than that of the negative current. However, when a negative ground fault occurs, the exact opposite will happen.

[0072] Therefore, the ratio of the two pole currents can be used as a fault selection criterion. The current ratio K is defined as follows:

[0073]

[0074] After a positive fault occurs on side S, the current ratio K will be much greater than 1 or much less than 1. Considering the necessary margin in the system and measurement errors, the fault pole selection criterion is set as follows:

[0075]

[0076] The number of sampling points for each comparison is N=10. A sliding time window is used to perform three consecutive calculations that satisfy the above criteria to determine the fault pole. Based on the fault pole selection criteria, the corresponding circuit breakers on both sides of line S and L trip, and the faulty line is disconnected.

[0077] Figure 1 This is a schematic diagram of a radial low-voltage DC distribution network topology with distributed generation. A transition resistance R is generated at the positive terminal of line 1. g Taking a single-pole ground fault as an example, in order to more intuitively analyze the changing pattern of the fault current, the system topology is reasonably simplified, and a series of lines and loads downstream of line 1 are transformed into an equivalent load, specifically as follows: Figure 2 The equivalent circuit model is shown.

[0078] Figure 3 This describes the fault detection method and IGBT latch-up logic. After a ground fault occurs, the grounding point and the VSC neutral point are connected through the earth, forming a fault loop. Therefore, the circuit of the VSC neutral point grounding branch will experience a sudden change after the fault. Even with a high transition resistance, the neutral point grounding branch current will still exhibit a significant sudden change. If the absolute value of this current change |ΔI| is detected... g If the value exceeds the start-up threshold, it can be determined that a ground fault has occurred. In this case, the IGBT in the system is locked out, the protection is activated, and the IGBT (G1) between the VSC three-phase bridge arm and the DC side output capacitor is turned off to cut off the connection between the AC and DC sides.

[0079] Figure 4This is the equivalent circuit diagram for a fault during the discharge phase of the VSC DC-side outlet capacitor. After a fault occurs, the VSC DC side can be equivalent to a second-order free-discharge capacitor circuit. In a DC distribution network, the equivalent resistance of a DC line is very small, and its damping generally satisfies... Therefore, its capacitor discharge can be regarded as a second-order underdamped oscillation process, and the positive fault current I p The current rises rapidly. As the positive capacitor continues to discharge, the rate of change of the positive current decreases continuously. Afterward, the capacitor discharge is essentially complete, and the capacitor's discharge rate is less than the inductor's discharge rate, causing the positive fault current i to rise rapidly. p It began to descend slowly.

[0080] Regarding the negative current, although the negative line is not directly connected to the fault circuit, the negative current still fluctuates. The change in the negative current (the same current flowing through the load) is closely related to the DC-side load voltage. Before the fault occurred, due to the small inductance L1, the load R... load Voltage U on load Approximately DC side voltage U dc In the initial stage of a fault, the load voltage U load The voltage drops to less than the non-faulty electrode voltage (DC side voltage U). dc The value of (half of) the negative electrode current I is therefore the negative electrode current I n It drops rapidly. Then, as the discharge continues, the rate of change of the positive electrode current, dI... p / d t The voltage continues to decrease, based on the voltage-current relationship U on inductor L1. L1 =L1(dI p / d t ), U L1 It also continues to decrease, the voltage drop across the non-faulty circuit increases, and the negative current I... n It begins to gradually recover. As the capacitor discharge continues, the load voltage will eventually decrease, and the negative fault current I... n It will also begin to decline slowly.

[0081] The side of the line closer to the neutral point of the VSC capacitor is designated as the S side, and the side farther from the neutral point is designated as the L side. From the above analysis, it can be seen that for the protection of the S side, when a forward fault occurs, the symmetrical structure of the positive and negative poles is disrupted, resulting in different current trends at the protection installation point. However, when a reverse fault occurs or the load fluctuates normally, the symmetrical structure remains intact, and the current trends at the protection installation point are essentially the same. The L side protection is the opposite, but it cannot distinguish between forward faults and load fluctuations; therefore, a blocking element needs to be added to the L side protection. Utilizing the characteristic that the S side protection can distinguish between forward faults and load fluctuations, the L side operation logic can be set as follows: the S side first determines the fault direction and transmits the result to the L side via communication. If the S side confirms that the fault occurred in the forward direction on its side, and the L side meets the forward direction criterion, then the fault can be considered to have occurred within the line.

[0082] Figure 5 To protect the overall action logic, it can be specifically divided into three parts: fault detection, fault location, and fault selection.

[0083] First, fault detection is performed using the neutral point grounding branch current of the VSC DC side output capacitor. When the absolute value of the current change exceeds the starting threshold, it can be determined that a ground fault has occurred, the protection is activated, the IGBT is blocked, and the connection between the AC side and the DC side of the system is cut off.

[0084] Afterwards, the positive and negative currents at each protection installation point in the system are sampled in real time, and the Pearson correlation coefficients of each side of each line are calculated. The algorithm calculation window is 20 sampling points. If the criteria are met three times consecutively under the sliding time window, it can be judged as a positive direction fault of the protection on that side. The communication process between S and L is as follows: (1) S side is judged according to the protection criteria of this side. If S side judges it as a positive direction fault, it sends an allow signal to L side and waits for the response from the other side; if the line is the end line, it can be directly judged as a fault in the line area based on the protection judgment result of S side. (2) After L side receives the protection signal, if it also judges it as a positive direction fault, it can be determined that the fault occurred in the line area and sends a response signal to S side at the same time; if it judges it as a reverse direction fault, it blocks the protection on this side. (3) After S side receives the confirmation signal from L side, it starts fault selection.

[0085] Finally, the current ratio K is calculated based on the current sampling data on side S. The number of sampling points N = 10 for each comparison is taken. The fault pole is determined if the pole selection criterion is met three consecutive times within the sliding time window. The corresponding circuit breakers on both sides of line S and L trip according to the fault pole selection criterion, and the faulty line is disconnected.

Claims

1. A method for protecting DC distribution network grounding faults based on the correlation between positive and negative pole currents, characterized in that: The steps are as follows: (1) Fault detection, the method of which is: capacitor neutral point grounding branch current I g A sudden change occurs after a fault, which can quickly and accurately reflect the fault and is used as a protection start criterion. After the fault is detected, the IGBT in the system is locked to realize the self-protection of VSC; the IGBT between the three-phase bridge arm of VSC and the DC side output capacitor is quickly turned off, cutting off the connection between the AC and DC sides. capacitor neutral point grounding branch current I g The activation criterion when a change occurs is shown in the following formula: ， in, This represents the absolute value of the sudden change in current in the neutral-point grounded branch of the capacitor. The threshold for activation should take into account the actual engineering situation and leave sufficient margin for protection; (2) Fault location, the steps are as follows: ① Sample the positive and negative currents at each protection installation point in the system, and calculate the Pearson correlation coefficient for each side of each line. The specific method is as follows: Use the Pearson correlation coefficient to reflect the correlation between the changing trends of the positive and negative currents on one side of the line. The Pearson correlation coefficient is defined as follows: ， in, i P = { x 1 ,x 2 ,…,x n } represents the positive current sampling value. i N = { y 1 ,y 2 ,…,y n } represents the negative electrode current sampling value. , The mean of the sampled data. m The two sides of the line are numbered, with the two sides being the S and L sides; ②The S side is judged according to the protection criteria. If the S side is judged to be a positive direction fault, an allow signal is sent to the L side and the response is waited for. If the line segment is the end line of the system, it can be directly judged as a fault in the line area based on the judgment result of the S side. ③ After receiving the protection signal, if the L side also meets the positive direction fault criteria, it can be determined that the fault is within the line area and an action signal is sent to the S side at the same time; if the L side does not meet the positive direction fault criteria, it is determined that the fault is outside the line area and the protection on this side is blocked. ④ After receiving the confirmation signal from the L side, the S side initiates fault polarity selection; (3) Fault polarity selection, the method of which is: Based on the characteristics of the fault current change on the S side of the line, after determining that the fault occurs in the positive direction, the positive and negative currents on the S side are used as the basis for pole selection. When a positive pole ground fault occurs, the positive current rises rapidly, while the negative current decreases slightly. The amplitude of the positive current will be much higher than that of the negative current. When a negative pole ground fault occurs, the opposite will happen. The corresponding circuit breakers on both sides of the line S and L trip according to the fault pole selection criteria, and the faulty line is disconnected. The fault selection uses the ratio of the two pole currents as the fault selection criterion, defining the current ratio. K As shown in the following formula: ， After a positive fault occurs on the S side, the current ratio K For values ​​much greater than 1 or much less than 1, considering the margin required in the system and measurement errors, the fault selection criterion is set as follows: 。 2. The DC distribution network grounding fault protection method based on the correlation between positive and negative pole currents according to claim 1, characterized in that: The Pearson correlation coefficient P m The value range is [-1, 1]. P m When =1, there is a perfect positive correlation; P m When the value is -1, there is a completely negative correlation; P m =0 indicates no correlation.

3. The DC distribution network grounding fault protection method based on the correlation of positive and negative pole currents according to claim 2, characterized in that: For S-side protection, considering the required margin of protection, the protection operation conditions can be set as follows: ； For L-side protection, a blocking element needs to be added, configured to only release the protection after receiving a permission signal from S-side protection. Therefore, the operating conditions for L-side protection can be set as follows: 。 4. The DC distribution network grounding fault protection method based on the correlation between positive and negative pole currents according to claim 1, characterized in that: Current ratio K Take the number of sampling points for each comparison N =10, and the fault pole can be determined by performing three consecutive calculations using a sliding time window that satisfy the fault pole selection criterion.

5. The DC distribution network grounding fault protection method based on the correlation between positive and negative pole currents according to claim 1, characterized in that: For the criteria of S-side protection and L-side protection, the calculation window of the Pearson correlation coefficient algorithm is 20 sampling points. If the criteria are met three times consecutively within the sliding window, it can be determined that the protection is in the positive direction and the protection will be activated.

Images