A Lightning Interference Identification Method for Flexible DC Transmission Lines Based on Pearson Coefficient

Abstract

This invention discloses a method for identifying lightning interference in flexible DC transmission lines based on the Pearson coefficient, specifically including the following steps: Step 1, collecting voltage and current data within a time window from 0.25ms to 1.25ms after the protection element starts, and calculating the line-mode current; Step 2, calculating the Pearson correlation coefficient γ between the line-mode current obtained in Step 1 and the corresponding time quantity; Step 3, if γ > γ set Then it is judged as a fault; if γ < γ set If this is the case, it is determined that lightning interference has occurred. This method can accurately distinguish between faults and lightning interference by obtaining the Pearson correlation coefficient between the line-mode current and time at the protection measuring point.

Description

Technical Field

[0001] This invention belongs to the field of power system technology and relates to a method for identifying lightning interference in flexible DC transmission lines based on the Pearson coefficient. Background Technology

[0002] Compared to traditional DC transmission technology, flexible DC transmission technology demonstrates significant advantages in renewable energy integration. It not only avoids commutation failure and reactive power compensation issues but also simultaneously provides active and reactive power to the AC system, thereby achieving efficient integration and complementarity of large-scale, multi-type renewable energy sources, improving energy utilization efficiency, and enhancing power supply stability. Because flexible DC transmission systems employ Modular Multilevel Converter (MMC) technology, their high degree of electronic integration leads to their designation as "low-inertia" systems. When a line fault occurs, overcurrent may occur within milliseconds. If the faulty line cannot be detected and isolated in a timely and accurate manner, the converter station may be shut down due to valve hall overcurrent protection, placing higher demands on the accuracy and response speed of the protection system.

[0003] In the line protection of flexible DC transmission systems, accurately and quickly isolating faulty lines is the core of the protection system. However, lightning interference is one of the most common external interferences in power systems. If a region is struck by lightning but no fault occurs, the protection device may malfunction. To ensure the correct operation of the protection device, it is necessary to conduct in-depth research on the transient traveling wave waveform characteristics of flexible DC transmission systems under different types of faults and lightning interference, and to propose a fast and reliable lightning interference identification standard.

[0004] Existing methods for identifying lightning interference primarily rely on the analysis and discrimination of traveling wave transients and their frequency domain characteristics. However, these methods still have numerous shortcomings and limitations in practical applications, affecting the accuracy and reliability of lightning interference identification.

[0005] First, traditional identification methods based on traveling waves and transient quantities are highly sensitive to high-frequency interference signals and are easily affected by lightning strikes, leading to malfunctions. For example, lightning interference signals typically contain a large number of high-frequency components, making it difficult for existing traveling wave protection devices to distinguish between lightning interference and actual faults, resulting in misjudgments. Second, while identification methods based on frequency domain characteristics can improve accuracy to some extent, they still face challenges in setting thresholds and exhibit poor noise robustness in complex electromagnetic environments. These methods require extensive parameter tuning and optimization in practical applications, increasing system complexity and maintenance costs.

[0006] Furthermore, existing lightning interference identification methods also have shortcomings in terms of speed and real-time performance. For highly sensitive applications such as flexible DC transmission systems, existing methods struggle to quickly determine the nature of lightning interference and short-circuit faults before the main protection system operates, thus affecting the system's protection performance. The identification performance of existing methods also needs further improvement under conditions of abnormal data transmission or noise interference.

[0007] In summary, existing lightning interference identification methods have obvious defects and limitations in terms of high-frequency interference sensitivity, threshold setting, identification speed, and real-time performance. Summary of the Invention

[0008] The purpose of this invention is to provide a method for identifying lightning interference in flexible DC transmission lines based on the Pearson coefficient. This method can accurately distinguish between faults and lightning interference by obtaining the Pearson correlation coefficient between the line mode current and time at the protection measuring point.

[0009] The technical solution adopted in this invention is a method for identifying lightning interference in flexible DC transmission lines based on the Pearson coefficient, which specifically includes the following steps:

[0010] Step 1: Collect data within the time window from 0.25ms after the protection element starts to 1.25ms after the protection element starts, and calculate the line mode current Δi1(k);

[0011] Step 2: Calculate the Pearson correlation coefficient γ between the line mode current obtained in Step 1 and the corresponding time quantity;

[0012] Step 3: Determine whether lightning interference has occurred based on the Pearson correlation coefficient γ calculated in Step 2.

[0013] The invention is further characterized by:

[0014] The specific process of step 1 is as follows:

[0015] Step 1.1: After the protection element is activated, set a protection time window. Within this time window, record the positive current i at the installation location of the protection device. p (k) and negative electrode current i n (k), and compared with the average positive current i within 10ms before the start of protection. p0 and the average value of the negative electrode current i n0 By subtracting the two values, the fault component Δi of the positive current can be calculated. p (k) and negative electrode current Δi n The fault component of (k) is calculated as shown in formula (1):

[0016]

[0017] Step 1.2, calculate the fault component Δi1(k) of the line-mode current within the time window. The calculation process is shown in equation (2):

[0018]

[0019] In step 1.1, the protection time window is set from 0.25ms to 1.25ms after the protection element is started.

[0020] The specific process of step 2 is as follows:

[0021] The Pearson correlation coefficient γ between the line-mode current fault component Δi1(k) and the time component is calculated according to the following formula (3):

[0022]

[0023] In the formula: k is the time sequence number, 0≤k≤N-1; N is the number of sampling points within the protection time window; This is the average current. This is the average of the sampled points.

[0024] The specific process of step 3 is as follows:

[0025] If γ>γ set Then it is judged as a fault; if γ < γ set This is then determined to be lightning interference, where γ set This is the threshold value.

[0026] The beneficial effects of this invention are that it identifies lightning interference based on the Pearson correlation coefficient between the single-ended line modulus current and the time vector. This method possesses good tolerance to transition resistance and noise interference. The identification method provided by this invention can still effectively identify lightning interference under 0Ω and 500Ω transition resistance conditions and 20dB noise interference. Attached Figure Description

[0027] Figure 1 This is a flowchart of the lightning interference identification method for flexible DC transmission lines based on Pearson coefficient according to the present invention;

[0028] Figure 2 This is a simulation model diagram of a true bipolar MMC flexible DC transmission system at both ends;

[0029] Figure 3 In the simulation verification stage, the lightning interference identification method of flexible DC transmission line based on Pearson coefficient of this invention was used to determine the judgment result of the local grounding fault at 250km of the bipolar flexible DC system with a transition resistance of 500Ω.

[0030] Figure 4 This is a simulation model diagram of a four-terminal MMC flexible DC transmission network;

[0031] Figure 5 In the simulation verification stage, the lightning interference identification method of flexible DC transmission lines based on Pearson coefficient of this invention was used to determine the lightning interference backlash identification at 100km of a four-terminal flexible DC power grid.

[0032] Figure 6 This is a simulation model diagram of a four-terminal mesh-type MMC flexible DC transmission network;

[0033] Figure 7 In the simulation verification stage, the lightning interference identification method of flexible DC transmission lines based on Pearson coefficient of this invention was used to determine the lightning fault bypass judgment at 100km in a four-terminal mesh flexible DC system. Detailed Implementation

[0034] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0035] This invention relates to a method for identifying lightning interference in flexible DC transmission lines based on the Pearson coefficient, which specifically includes the following steps:

[0036] Step 1: Collect data within the time window from 0.25ms after the protection element starts to 1.25ms after the protection element starts, and calculate the line mode current Δi1(k);

[0037] The specific process of step 1 is as follows: Step 1.1, after the protection element is activated, a protection time window is set from 0.25ms to 1.25ms after the protection element is activated. Within this time window, the positive current i at the installation location of the protection device is recorded. p (k) and negative electrode current i n (k), and compared with the average positive current i within 10ms before the start of protection. p0 and the average value of the negative electrode current i n0 By subtracting the two values, the fault component Δi of the positive current can be calculated. p (k) and negative electrode current Δi n The fault component of (k). The calculation process is shown in formula (1):

[0038]

[0039] Step 1.2, calculate the fault component Δi1(k) of the line-mode current within the time window. The calculation process is shown in equation (2):

[0040]

[0041] Step 2: Calculate the Pearson correlation coefficient γ between the line mode current obtained in Step 1 and the corresponding time quantity;

[0042] The specific process of step 2 is as follows:

[0043] The Pearson correlation coefficient between the line-mode current fault component Δi1(k) and the time component is calculated according to the following formula (3):

[0044]

[0045] In the formula: k is the time sequence number, 0≤k≤N-1; N is the number of sampling points within the protection time window; is the average current; k is the average of the sampling points.

[0046] Step 3, if γ > γ set Then it is judged as a fault; if γ < γ set This indicates that a lightning strike has occurred.

[0047] The specific process of step 3 is as follows: If γ > γ set Then it is judged as a fault; if γ < γ set If this is the case, it is determined that lightning interference has occurred. The line-mode current waveform within the time window used in this invention shows a curve positively correlated with the time vector under fault conditions; and a curve negatively correlated or not positively correlated with the time vector under lightning interference conditions. Therefore, this invention considers that the measured γ is close to 1 under fault conditions, while the measured γ is close to 0 or -1 under lightning interference conditions. Based on simulation data, a threshold value γ is determined. set =0.5.

[0048] As attached Figure 2 The figure shows a simulation model of a double-ended flexible DC transmission system. In the figure, the current-limiting reactor L = 0.2H, MMC1 and MMC2 are modular converters, P1 and P2 are line boundary protection measurement points, and f1 is the location of the disturbance. This system has a rated voltage of ±500kV, a rated transmission capacity of 3000MVA, and a total transmission line length of 500km. An overhead line frequency-varying parameter model is used, and its line surge impedances are Z... c0 =320Ω, Z c1 =260Ω. During system simulation, the sampling frequency was 100kHz, and a 1.2 / 50μs standard lightning current model was used to simulate a lightning strike. The entire text uses Line 1, measuring point P1 as an example; the fault distance is the distance from the fault location to measuring point P1, as shown in the attached figure. Figure 1 The identification process identifies grounding faults.

[0049] Example 1

[0050] like Figure 3 As shown, in Figure 2The simulated waveform of the line-mode current measured at the protection measuring point P1 in the dual-ended flexible DC transmission system shown is obtained when a positive ground fault with a transition resistance of 500Ω occurs 250km away from the protection measuring point P1. After the protection element is activated, the line-mode current data from the 25th sampling point to the 125th sampling point after activation are read, and the Pearson correlation coefficient γ between the line-mode current and the time sampling points within the data window is calculated. γ > γ set The system was flagged as a fault and activated its protection mechanism.

[0051] As attached Figure 4 The figure shows a simulation model of a four-terminal MMC flexible DC transmission network. In the figure, the current-limiting reactor L = 0.2H, MMC1, MMC2, MMC3, and MMC4 are modular converters, P1 and P2 are line boundary protection measurement points, f1 is the location of the disturbance, the rated voltage between MMC1 and MMC2 is ±500kV, the rated transmission capacity is 3000MVA, the total length of the transmission line is 207.9km, and an overhead line frequency-varying parameter model is used. The line surge impedances are Z... c0 =310Ω, Z c1 =250Ω. During system simulation, the sampling frequency is 100kHz, and a 1.2 / 50μs standard lightning current model is used to simulate lightning strikes. Because the four-terminal MMC flexible DC network is highly symmetrical except for the MMC parameters, this invention only takes the protection measuring point P1 of the protected line Line1 as an example to analyze the lightning interference identification method. The fault distance is the distance from the fault location to measuring point P1, according to the attached... Figure 1 The identification process identifies lightning interference.

[0052] Example 2

[0053] As attached Figure 5 As shown, in Figure 4 The waveform of the line-mode current measured at the protection measuring point is shown when a lightning strike occurs 100km away from the protection installation location in a four-terminal MMC flexible DC power grid. After the protection element is activated, the line-mode current data from the 25th sampling point to the 125th sampling point after activation are read, and the Pearson correlation coefficient γ between the line-mode current and the time sampling points within the data window is calculated, where γ < γ set The incident was determined to be caused by lightning interference, and the system was activated and returned to normal.

[0054] As attached Figure 6 The figure shows a simulation model of a four-terminal mesh-type MMC flexible DC transmission network. In the figure, the current-limiting reactor L = 0.2H, MMC1, MMC2, MMC3, and MMC4 are modular converters, P1 and P2 are line boundary protection measurement points, f1 is the location of the disturbance, the rated voltage between MMC1 and MMC2 is ±500kV, the rated transmission capacity is 3000MVA, the total length of the transmission line is 207.9km, and an overhead line frequency-varying parameter model is used. The line surge impedances are Z...c0 =310Ω, Z c1 =250Ω. During system simulation, the sampling frequency was 100kHz, and a 1.2 / 50μs standard lightning current model was used to simulate lightning strikes. Because the four-terminal "mesh-type" flexible DC network is highly symmetrical except for the MMC parameters, this paper only takes the protection measuring point P1 of the protected line Line1 as an example to analyze the lightning interference identification method. The fault distance is the distance from the fault location to measuring point P1, as shown in the attached... Figure 1 The identification process identifies lightning strike faults.

[0055] Example 3

[0056] like Figure 7 As shown, in Figure 6 When a lightning strike fault occurs 100km from the protection installation point in a four-terminal "mesh-type" flexible DC power grid, the line-mode current waveform measured at the protection measuring point is recorded. After the protection element is activated, line-mode current data from the 25th sampling point to the 125th sampling point after activation are read. The Pearson correlation coefficient γ between the line-mode current within the data window and the time sampling points is calculated, where γ > γ set The system was flagged as a fault and activated its protection mechanism.

[0057] To comprehensively verify the impact of fault distance, lightning strike type, transition resistance magnitude, and their influence on the discrimination results, a bipolar DC transmission system was set up with lightning faults, lightning interference, and grounding faults with transition resistances of 0Ω and 500Ω at distances of 0km, 100km, 250km, 400km, and 500km, respectively. The influence of 20dB noise was also considered, and the lightning interference identification method was verified based on the simulation results.

[0058] Table 1 shows the simulation data for different fault distances on the bipolar flexible DC transmission line. From the data within the time window, we can see that the Pearson correlation coefficient is less than 0.5 when lightning interference occurs; and greater than 0.5 when both lightning and grounding faults occur. Lightning interference can be identified at all fault distances.

[0059] Table 1 Simulation verification results for different fault distances

[0060]

[0061] Table 2 provides simulation data for ground faults with different transition resistances on bipolar flexible DC transmission lines. The data within the time window shows that the Pearson correlation coefficient is greater than 0.5 when a ground fault occurs, indicating that the fault can be accurately identified under different transition resistance ground faults.

[0062] Table 2 Simulation verification results for different transition resistances

[0063]

[0064] Table 3 provides simulation data for lightning strikes and backflashovers on bipolar flexible DC transmission lines. Data within the time window shows that the Pearson correlation coefficient is less than 0.5 when lightning interference occurs. Both lightning strikes and backflashovers can be accurately identified.

[0065] Table 3 Simulation verification results for different lightning strike types

[0066]

[0067] Table 4 provides simulation data for faults in bipolar flexible DC transmission lines under 20dB noise interference. Data within the time window shows that the Pearson correlation coefficient is less than 0.5 when lightning interference occurs; and greater than 0.5 when both lightning and grounding faults occur. The protection criterion can accurately identify lightning interference under 20dB noise interference.

[0068] Table 4 Simulation verification results under the influence of 20dB noise interference

[0069]

[0070]

Claims

1. A method for identifying lightning interference in flexible DC transmission lines based on Pearson coefficient, characterized in that: Specifically, the steps include the following: Step 1: Collect data within a time window from 0.25 ms to 1.25 ms after the protection element starts, and calculate the line-mode current. The specific process of step 1 is as follows: Step 1.1: After the protection element is activated, set a protection time window. Within this time window, record the positive current at the installation location of the protection device. and negative current And compared with the average positive current within 10 ms before the protection is activated. and the average value of the negative electrode current The fault component of the positive current is calculated by subtracting the two. and negative current The fault component is calculated as shown in formula (1): （1） In step 1.1, a protection time window is set from 0.25 ms to 1.25 ms after the protection element is activated; Step 1.2: Calculate the fault component of the line-mode current within the time window. The calculation process is shown in equation (2): （2） Step 2: Calculate the Pearson correlation coefficient between the line-mode current obtained in Step 1 and the corresponding time quantity. The specific process of step 2 is as follows: calculate the fault component of the line-mode current according to the following formula (3). Pearson correlation coefficient with time quantity : （3） In the formula: k For time sequence number, 0≤ k ≤ N -1; N To protect the number of sampling points within the time window; This is the average current. The average of the sampled points; Step 3, based on the Pearson correlation coefficient calculated in Step 2 To determine whether lightning interference has occurred, step 3 specifically involves the following process: If > Then it is judged as a fault; if < This is then determined to be interference caused by a lightning strike. This is the threshold value.

Images