New energy power transmission line dynamic mode identification protection method, system, device and medium

Abstract

The invention relates to the technical field of power system relay protection, in particular to a new energy power transmission line dynamic mode recognition protection method, system and device and a medium. Whether a protection algorithm is started or not is judged through the phase current break variable; acquiring a three-phase current sampling value at a protection installation position, and sending current sampling data of a home terminal to an opposite terminal protection device; after current sampling data of an opposite-end protection device are obtained, segmentation mode division is carried out on the current sampling sequence data; combining the mean value information of the segmented current sequence with the change direction to generate an improved mode symbol based on a ternary mode; calculating a dynamic mode matching distance between the currents at the two ends; and comparing the dynamic mode matching distance with a preset setting value, if the dynamic mode matching distance is greater than the setting value, judging that an internal short-circuit fault occurs and sending a tripping command, otherwise, judging that normal operation or external short-circuit fault occurs.

Description

Technical Field

[0001] The present invention relates to the technical field of power system relay protection, and in particular to a dynamic pattern recognition protection method, system, equipment and medium for a new energy power transmission line. Background Art

[0002] As global demand for clean energy continues to grow, China is increasing its investment and development in the new energy sector. Policy support and technological advancements have driven the rapid expansion of renewable energy sources such as wind and solar power. The integration of large-scale photovoltaics has altered the power flow distribution and fault current characteristics of the power grid, placing higher demands on grid protection.

[0003] Traditional grid protection methods face numerous challenges after the integration of renewable energy stations. Pilot protection methods based on cosine similarity and person coefficients offer good performance under rated operating conditions for renewable energy stations, but neither approach addresses computational challenges in weak output scenarios. Pilot protection methods based on improved Euclidean distance can effectively address the changing characteristics of faults after renewable energy integration, but they can only handle time series of equal length and cannot identify changing trends.

[0004] Existing longitudinal protection methods based on Tanimoto similarity, while suitable for weak PV station output scenarios, suffer from high computational complexity. Waveform similarity criteria based on dynamic time warping, while supporting timeline expansion and curvature and effectively measuring the similarity of sequences of unequal lengths, require the calculation of a cost matrix based on Euclidean distance and its optimal curving path. These methods are sensitive to numerical variations, have high computational and time complexity, and are unable to meet the requirements for rapid protection. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention is proposed.

[0006] Therefore, the problem to be solved by the present invention is how to solve the problem that the protection performance of the existing new energy AC line current longitudinal protection method is easily reduced under complex fault characteristics and current data environments, as well as the technical problems that the traditional protection method has high calculation complexity, slow response speed, and insufficient reliability under various operating conditions of new energy stations.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In a first aspect, an embodiment of the present invention provides a dynamic pattern recognition protection method for a new energy transmission line, which includes determining whether to start a protection algorithm based on a phase current mutation amount;

[0009] Obtain the three-phase current sampling values ​​at the protection installation location and send the local current sampling data to the opposite protection device;

[0010] After obtaining the current sampling data of the opposite end protection device, the current sampling sequence data is divided into segment modes;

[0011] The mean information of the segmented current sequence is combined with the change direction to generate an improved pattern symbol based on the ternary pattern;

[0012] Calculate the dynamic mode matching distance between the currents at both ends;

[0013] The dynamic mode matching distance is compared with the preset setting value. If it is greater than the setting value, it is judged that an internal short circuit fault has occurred and a trip command is sent. Otherwise, it is judged to be normal operation or an external short circuit fault.

[0014] As a preferred solution of the dynamic pattern recognition protection method for new energy transmission lines of the present invention, the method of judging whether to start the protection algorithm based on the phase current mutation amount includes:

[0015] Calculate the phase current mutation in real time; start the protection algorithm when the mutation of any phase current exceeds 0.1 times the load current;

[0016] The protection starts when three consecutive sampling points meet the start criteria, otherwise the protection is reset.

[0017] As a preferred solution of the dynamic pattern recognition protection method for new energy transmission lines of the present invention, the step of generating an improved pattern symbol based on a ternary pattern includes:

[0018] The pattern symbol type is determined according to the mean value and slope positivity of the current sequence;

[0019] The three states of rising, falling and holding correspond to different mode symbols respectively;

[0020] The symbols of the hold states under different mean values ​​are uniformly set to the same identifier to improve computational efficiency.

[0021] As a preferred solution of the dynamic pattern recognition protection method for new energy transmission lines of the present invention, the calculation of the dynamic pattern matching distance between the currents at both ends includes:

[0022] After converting the current sampling data into pattern symbols, the pattern matching distance is calculated; a morphological pattern matrix between two sets of morphological pattern sequences is constructed;

[0023] A dynamic programming algorithm is used to calculate the cumulative distance matrix; the dynamic pattern matching distance is determined by the final value of the cumulative distance matrix.

[0024] As a preferred solution of the dynamic pattern recognition protection method for new energy transmission lines of the present invention, the preset setting value is determined according to the number of segments of the current sequence:

[0025] According to the current sequence data extracted from the 10ms data window, the number of segments is determined to be 100; the dynamic mode matching setting value is set to 30;

[0026] When the dynamic mode matching distance is greater than 30, it is determined that the phase current has a fault.

[0027] As a preferred solution of the dynamic pattern recognition protection method for new energy transmission lines of the present invention, the identification of internal and external faults includes:

[0028] Calculate the dynamic pattern matching distance of the three-phase current respectively; compare the dynamic pattern matching distance of each phase with the setting value 30; identify the specific fault phase and output the corresponding protection action.

[0029] As a preferred solution of the dynamic pattern recognition and protection method for new energy transmission lines described in the present invention, the specific fault phases include the ability to identify single-phase grounding faults, two-phase short circuit faults, two-phase grounding short circuit faults and three-phase short circuit faults, and maintain recognition accuracy under different signal-to-noise ratios and high transition resistance conditions.

[0030] In a second aspect, an embodiment of the present invention provides a dynamic pattern recognition protection system for a new energy transmission line, which includes a protection algorithm startup unit for acquiring three-phase currents at a protection installation in real time, calculating a phase current mutation amount, and starting a protection algorithm when the mutation amount exceeds a threshold;

[0031] A communication unit, used for sending current sampling data to a protection device at the opposite end;

[0032] A dynamic pattern matching distance calculation unit is used to obtain current sampling data of the opposite side protection device and calculate the dynamic pattern matching distance;

[0033] The fault judgment unit is used to compare the size relationship between the action amount and the set value, judge the fault inside and outside the zone and output the protection action instruction.

[0034] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the dynamic pattern recognition and protection method for new energy transmission lines as described in the first aspect of the present invention are implemented.

[0035] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored, wherein: when the computer program instructions are executed by a processor, the steps of the dynamic pattern recognition and protection method for new energy transmission lines as described in the first aspect of the present invention are implemented.

[0036] The present invention has the following beneficial effects: Through its innovative dynamic pattern matching algorithm, it effectively addresses the technical challenge of reduced performance of traditional renewable energy AC line protection methods in complex fault environments. Compared to traditional cosine similarity and Pearson coefficient methods, this method avoids the zero-division-by-zero calculation problem in weak renewable energy station output scenarios, significantly improving the numerical stability and computational reliability of the protection algorithm.

[0037] This method uses an improved symbol generation technique based on ternary patterns to transform complex current waveforms into simplified pattern sequences, preserving the waveform's key characteristics while significantly reducing computational complexity. Compared to traditional Euclidean distance methods, this approach not only handles sequences of equal length but also effectively addresses the issue of sequence length discrepancies caused by asynchronous sampling during actual operation, enhancing the algorithm's adaptability and practicality.

[0038] Compared to dynamic time warping methods, this method avoids the complex cost matrix calculation based on Euclidean distance by simplifying the pattern symbols, significantly reducing time complexity and meeting the stringent requirements of power system protection for fast response. Furthermore, this method significantly reduces sensitivity to numerical changes, improving its ability to resist interference in noisy environments.

[0039] This invention boasts excellent fault identification capabilities and environmental adaptability. Through a phase current mutation triggering mechanism and dynamic pattern matching distance calculation, it can accurately identify a variety of fault types, including single-phase ground faults, two-phase short circuit faults, two-phase ground short circuit faults, and three-phase short circuit faults. It maintains stable identification accuracy under varying signal-to-noise ratios and high transition resistance conditions, demonstrating excellent robustness.

[0040] The protection scheme of this invention ensures both reliability and selectivity while meeting rapidity requirements, providing a strong guarantee for the safe and stable operation of new energy stations. The entire technical solution, through collaborative software and hardware design, achieves the integrated integration of data acquisition, communication transmission, algorithm calculation, and protection actions, demonstrating excellent engineering practical value and promising prospects for widespread application. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 A flow chart of the dynamic pattern recognition protection method for renewable energy transmission lines;

[0043] Figure 2Computer equipment diagram for dynamic pattern recognition protection method of new energy transmission lines;

[0044] Figure 3 Schematic diagram of the AC line simulation model when a grid-forming (GFM) photovoltaic inverter is connected to the dynamic pattern recognition protection method for renewable energy transmission lines;

[0045] Figure 4 The dynamic pattern recognition protection method for new energy transmission lines is used to identify the AC line Figure 3 Schematic diagram of system AC line fault when a ground fault occurs at point F1;

[0046] Figure 5 Dynamic pattern recognition protection method for new energy transmission lines Figure 3 The schematic diagram of the fault current waveform after a three-phase short circuit fault is set at point F1;

[0047] Figure 6 Another flow chart of the new energy transmission line protection method of the dynamic pattern recognition protection method for new energy transmission lines. DETAILED DESCRIPTION

[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0049] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0050] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it constitute a separate or selective embodiment that is mutually exclusive with other embodiments.

[0051] Example 1

[0052] Reference Figures 1 and 2 , which is the first embodiment of the present invention, provides a new energy transmission line dynamic pattern recognition protection method, including:

[0053] S100: Determine whether to start the protection algorithm based on the phase current mutation amount;

[0054] S200: Obtain three-phase current sampling values ​​at the protection installation location and send the current sampling data at the local end to the protection device at the opposite end;

[0055] S300: After obtaining the current sampling data of the opposite protection device, the current sampling sequence data is divided into segment modes;

[0056] S400: combining the mean value information of the segmented current sequence with the change direction to generate an improved pattern symbol based on the ternary pattern;

[0057] S500: Calculating a dynamic mode matching distance between the currents at both ends;

[0058] S600: Compare the dynamic mode matching distance with the preset setting value. If it is greater than the setting value, it is determined that an internal short circuit fault has occurred and a trip command is sent. Otherwise, it is determined to be normal operation or an external short circuit fault.

[0059] In actual operation, the renewable energy transmission lines are faced with complex fault characteristics and changeable operating conditions. After the traditional current longitudinal protection method is connected to the renewable energy station, the control characteristics and output characteristics of the photovoltaic inverter are significantly different from those of the traditional synchronous generator, resulting in changes in the amplitude, phase and frequency characteristics of the fault current, causing the protection method based on the traditional similarity algorithm to malfunction or refuse to operate. Specifically, when the renewable energy station is in a weak output state, the fault current is small and the waveform is severely distorted. The traditional cosine similarity and Pearson coefficient methods are prone to 0 / 0 calculation problems; and the method based on Euclidean distance can handle numerical changes, but can only handle sequences of equal length and cannot adapt to the situation of asynchronous sampling in actual operation; the method based on dynamic time bending supports sequence length differences, but the computational complexity is too high to meet the rapidity requirements of protection.

[0060] Through the complete technical solution of S100-S600, the present invention constructs a new energy transmission line protection system based on dynamic pattern matching. First, the phase current mutation startup mechanism (S100) is used to ensure that the protection algorithm can respond promptly when a fault occurs, avoiding false operation during normal operation; then, complete current information is obtained through two-terminal communication (S200), providing a data basis for subsequent analysis; then, through segmented pattern division (S300) and ternary pattern symbol generation (S400), the complex current waveform is converted into a simplified pattern sequence, which not only retains the key features of the waveform but also reduces the computational complexity; then, through dynamic pattern matching distance calculation (S500), accurate quantification of the similarity of the currents at both ends is achieved. This method not only avoids the numerical sensitivity problem of traditional methods, but also can handle sequences of unequal lengths; finally, through comparison with the preset setting value (S600), accurate identification of faults inside and outside the zone and output of corresponding protection actions are achieved. The entire solution not only ensures the reliability and selectivity of protection, but also meets the requirements of rapidity.

[0061] Example 2

[0062] Reference Figure 2-Figure 6 , which is the second embodiment of the present invention.

[0063] In the embodiment of the present application, determining whether to start the protection algorithm based on the phase current sudden change in step S100 includes the following steps A1-A2:

[0064] A1: Determine whether to start the protection algorithm based on the phase current mutation, including:

[0065] Calculate the phase current mutation in real time; start the protection algorithm when the mutation of any phase current exceeds 0.1 times the load current;

[0066] The protection starts when three consecutive sampling points meet the start criteria, otherwise the protection is reset.

[0067] Determine whether to start the protection algorithm based on the phase current mutation amount;

[0068] If the sudden change trigger is met, the three-phase current sampling value at the protection installation is obtained and the local current sampling data is sent to the protection device at the opposite end;

[0069] After obtaining the current sampling data of the protection device at the opposite end, the dynamic mode matching distance between the currents at the two ends is calculated;

[0070] like It is determined that a short circuit fault has occurred in the area and a trip command is sent;

[0071] Otherwise, it is judged as During normal operation or an out-of-zone short circuit fault, the protection device is reset.

[0072] The specific implementation of real-time calculation of phase current mutation includes continuous sampling and differential calculation of three-phase current. The system collects the A, B, and C three-phase currents at a sampling frequency of 50kHz in real time, and calculates the difference between the current sampling value and the average value in the previous cycle in each sampling cycle. The mutation calculation adopts a sliding window mechanism, and the window length is set to the number of sampling points N corresponding to one power frequency cycle. When any phase current When the difference between the current value and the historical average value exceeds 0.1 times the load current, the system determines that a sudden current change has occurred. This design effectively distinguishes between normal load fluctuations and current changes caused by faults, avoiding protection misoperation caused by normal operations such as load switching.

[0073] Whether to start protection is determined by the phase current mutation amount, specifically:

[0074] The phase current mutation is calculated in real time. When the mutation of any phase current exceeds 0.1 times the load current, the protection algorithm is activated. The expression is as follows:

[0075]

[0076] Where: i φ (n) represents the value of the nth sampling point of a phase current; Indicates phase A, phase B or phase C; N is the number of sampling points in one cycle; I set is the set current value, which is 0.1 to 0.3 times the load current value. N is the number of sampling points in a cycle. When three consecutive sampling points meet any of the criteria in the above formula, the protection is activated; otherwise, the protection is reset.

[0077] A2: The continuity verification mechanism for the triggering criteria ensures the accuracy of protection action. The system requires that the mutation criterion be met at three consecutive sampling points before the protection triggering condition is considered met. This design effectively filters out false triggers caused by measurement noise or transient interference. Once the triggering condition is met, the protection algorithm enters the operating state and begins the subsequent data collection and analysis process. If the mutation criterion returns to normal during subsequent sampling, the protection algorithm automatically reverts to the monitoring state.

[0078] In an optional embodiment, the mutation amount calculation in step S100 can also adopt an adaptive threshold setting method to dynamically adjust the start threshold according to the current system operating status and historical data to adapt to the impact of changes in the output of new energy stations on protection sensitivity.

[0079] It should be noted that the design of the phase current mutation starting mechanism fully considers the operating characteristics of new energy stations. It can not only respond to various faults in a timely manner, but also avoid unnecessary actions caused by fluctuations in new energy output.

[0080] In the embodiment of the present application, obtaining the three-phase current sampling values ​​at the protection installation and performing data communication in step S200 includes the following steps B1-B2:

[0081] B1: Three-phase current sampling is acquired using high-precision current transformers and synchronous sampling technology. The protection device acquires the three-phase current signals at both ends of the line via dedicated current transformers (CTs). The sampling frequency is set to 50kHz to accurately capture the current variation characteristics during fault transients. The sampled data undergoes digital filtering and per-unit normalization to form standardized current sequence data. The system uses a GPS-synchronized clock to ensure time synchronization between sampling at both ends, with synchronization accuracy reaching microseconds.

[0082] If the sudden change trigger is met, the three-phase current sampling values ​​at the protection installation are obtained and the local current sampling data is sent to the protection device at the opposite end.

[0083] After obtaining the current sampling data of the opposite-end protection device, the dynamic mode matching distance between the currents at both ends is calculated.

[0084] After obtaining the current sampling data of the opposite protection device, the dynamic mode matching distance between the two-end currents is calculated. The expression is as follows:

[0085]

[0086] Among them I wφ and I sφ are the φ phase sampling currents on the photovoltaic side and the system side respectively. r(m,n) is the final value of the accumulated distance matrix L. L={r i,j} m×n , r i,j =D(i,j)+min{r i-1j ,r i,j-1 ,r i-1,j-1}. In the formula, r i,j is the cumulative distance, r(0,0)=0, r(i,0)=r(0,j)=∞.

[0087] Before calculating the dynamic pattern matching distance, the current sampling sequence data must be segmented into patterns. The mean information of the segmented current sequence must be combined with the direction of change. This allows the resulting pattern symbols to reflect both local morphology and overall trends. The improved pattern symbols based on the ternary model are shown in Table 1.

[0088] Table 1 Improved pattern symbols based on ternary patterns

[0089]

[0090] In order to improve the computational efficiency, the symbols of the hold states under different mean values ​​are all set to E. represents the mean value of the current sequence in the i-th segment, k i Represents the slope of the i-th segment fitting curve. min and x max Represent the minimum and maximum values ​​in the current sequence, that is, x min =min{x1,x2,x3…x n}、x max =max{x1,x2,x3…x n},

[0091] The pattern matching distance of the current sequence refers to the distance between the two patterns after the current sampling data is converted into segmented patterns, that is:

[0092]

[0093] where z i and z jThe pattern matching distance is actually the comparison of the similarities and differences between the pattern symbols of the current sampling segment subsequences. If the symbols are the same, the distance is 0, and if the symbols are different, the distance is 1.

[0094] Through the pattern matching distance, the morphological pattern matrix D between two sets of morphological pattern sequences can be constructed m×n

[0095] Each element D(i,j) of the matrix represents the matching distance between the i-th pattern and the j-th pattern in the two morphological pattern sequences. Then, according to the solution process of the dynamic time warping (DTW) distance, the dynamic programming algorithm is used to calculate the distance between the i-th pattern and the j-th pattern in D m×n Based on the calculation of the cumulative distance matrix L = {r i,j} m×n , r i,j =D(i,j)+min{r i-1j ,r i,j-1 ,r i-1,j-1}. In the formula, r i,j is the cumulative distance, r(0,0)=0, r(i,0)=r(0,j)=∞. The dynamic pattern matching distance between the two current sequences can be described by the final value rm,n of the cumulative distance matrix L, that is,

[0096] DPM(I wφ ,-I sφ )=r(m,n)

[0097] Among them I wφ and I sφ They are the φ-phase sampling currents on the photovoltaic side and the system side respectively.

[0098] B2: A dedicated protection communication channel is used to transmit current sampling data to the peer protection device. The system transmits local current sampling data to the peer protection device in real time via a fiber-optic communication network, with transmission delays of less than 2ms. Data transmission utilizes a dedicated protection communication protocol, including data integrity checks and retransmission mechanisms, ensuring data reliability. The system also monitors communication status in real time and automatically switches to local protection mode if communication is interrupted.

[0099] In an optional embodiment, the data communication in step S200 may also adopt a redundant communication channel design to further improve communication reliability through a primary and backup channel, and automatically switch to the backup channel when the primary channel fails.

[0100] It should be noted that high-quality current sampling and reliable data communication are the basis for achieving precise protection. This solution ensures the accuracy and real-time performance of data through the collaborative design of hardware and software.

[0101] In the embodiment of the present application, the current sampling sequence data is divided into segmented modes in step S300, including the following steps C1-C2:

[0102] C1: Determine the segmentation strategy and number of segments. Based on the temporal characteristics and sampling frequency of the fault transient process, the system divides the current sequence data within a 10ms time window into 100 subsegments, each containing approximately 5 sampling points. This segmentation approach maintains sufficient temporal resolution to capture fault characteristics while simplifying complex waveforms into manageable pattern sequences. Segment boundaries utilize sliding window technology to ensure continuity between adjacent segments.

[0103] C2: Establishing statistical characteristic parameters within each segment. Statistical parameters such as mean, standard deviation, and trend slope are calculated for the current data within each segment. The combination of these parameters accurately describes the current variation characteristics within that segment. The system also calculates inter-segment connectivity characteristics to describe the overall variation trend of the current waveform. These characteristic parameters provide the data foundation for subsequent pattern symbol generation.

[0104] In an optional embodiment, the segmentation in step S300 can also adopt an adaptive segmentation method to dynamically adjust the number of segments according to the complexity of the current waveform, increase the segment density for areas with drastic waveform changes, and reduce the number of segments for relatively flat areas.

[0105] It should be noted that a reasonable segmentation strategy is the key to achieving effective pattern recognition, which requires retaining sufficient detail information while avoiding excessive complexity that affects computational efficiency.

[0106] In the embodiment of the present application, generating an improved pattern symbol based on a ternary pattern in step S400 includes the following steps D1-D3:

[0107] D1: Generates improved pattern notation based on ternary patterns, including:

[0108] The pattern symbol type is determined according to the mean value and slope positivity of the current sequence;

[0109] The three states of rising, falling and holding correspond to different mode symbols respectively;

[0110] The symbols of the hold states under different mean values ​​are uniformly set to the same identifier to improve computational efficiency.

[0111] Determine the basic symbol type of the ternary mode. The system defines seven basic mode symbols: A, B, C, D, E, F, and G, which correspond to different current change mode combinations. Figure 1The improved symbol table based on the ternary pattern is shown in the figure, where E represents the hold state, A, B, C, and D represent the rising and falling states under different mean conditions, respectively, and F and G represent special change patterns. This symbol design can accurately describe the changing characteristics of the current series in different time periods.

[0112] D2: Combine the mean information and the change direction to generate the pattern symbol. For each segment, the system first calculates the current mean μ of the segment. i , and then calculate the slope k of the fitting curve i The symbol of the pattern corresponding to the segment is determined based on the size relationship of the mean relative to the mean of the entire series (greater than, equal to, less than) and the positive or negative slope. This design takes into account both local variation characteristics and overall trend information.

[0113] D3: Optimizing the handling of hold state symbols. To improve computational efficiency, the system uniformly represents hold states (i.e., those with slopes close to zero) under different mean values ​​using the symbol E. This simplified processing maintains pattern recognition accuracy while significantly reducing the complexity of subsequent matching calculations.

[0114] In an optional implementation, the pattern symbol generation in step S400 may also introduce a self-learning mechanism to continuously optimize the symbol classification rules by analyzing historical fault cases to improve the accuracy of pattern recognition.

[0115] It should be noted that the improved symbol design based on the ternary model fully considers the complexity of the current characteristics of new energy stations and realizes the effective description of complex waveforms through a limited set of symbols.

[0116] In the embodiment of the present application, calculating the dynamic mode matching distance between the currents at both ends in step S500 includes the following steps E1-E4:

[0117] E1: Calculate the dynamic mode matching distance between the currents at both ends, including:

[0118] After converting the current sampling data into pattern symbols, the pattern matching distance is calculated; a morphological pattern matrix between two sets of morphological pattern sequences is constructed;

[0119] A dynamic programming algorithm is used to calculate the cumulative distance matrix; the dynamic pattern matching distance is determined by the final value of the cumulative distance matrix.

[0120] Construct a pattern matching distance matrix. After converting the current sequences at both ends into a sequence of pattern symbols, calculate the matching distance between the symbols at corresponding positions. The matching distance for identical symbols is 0, and the matching distance for different symbols is 1. This simplified distance definition maintains computational simplicity while effectively distinguishing different pattern features.

[0121] E2: Establish the morphological pattern matrix M. Each element M of this matrix ij It represents the matching distance between the i-th pattern and the j-th pattern in two morphological pattern sequences. The dimension of the matrix is ​​determined by the number of segments of the current sequence at both ends and is usually a 100×100 matrix structure.

[0122] E3: Use a dynamic programming algorithm to calculate the cumulative distance matrix L. Based on the calculation principle of dynamic time warping (DTW), the system uses a dynamic programming method to calculate the cumulative distance matrix. The cumulative distance of the optimal path is calculated recursively.

[0123] E4: Determine the final dynamic pattern matching distance DPM. By accumulating the final value r of the distance matrix L m,n To characterize the dynamic mode matching distance between two current sequences, the specific calculation method is performed according to the DPM calculation formula.

[0124] In an optional implementation, the distance calculation in step S500 may further introduce a weight coefficient to assign different weights to pattern matching in different time periods, so as to highlight characteristic information at the early stage of the fault.

[0125] It should be noted that the calculation of dynamic pattern matching distance fully draws on the advantages of the DTW algorithm, while avoiding the problem of high computational complexity of the traditional DTW method by simplifying the pattern symbols.

[0126] In the embodiment of the present application, the fault diagnosis and protection action are performed in step S600, including the following steps F1-F:3:

[0127] F1: The preset setting value is determined by the number of segments of the current sequence:

[0128] According to the current sequence data extracted from the 10ms data window, the number of segments is determined to be 100; the dynamic mode matching setting value is set to 30;

[0129] When the dynamic mode matching distance is greater than 30, it is determined that the phase current has a fault.

[0130] Set the dynamic mode matching setting value. According to the current sequence data extracted from the 10ms data window, when the number of segments is 100, the system will set the dynamic mode matching setting value DPM set Set to 30. This setting, determined through extensive simulations and field tests, effectively distinguishes between normal operation, out-of-zone faults, and in-zone faults. During normal operation or an out-of-zone short circuit, the dynamic mode matching distance between the currents at both ends is typically less than 30, while during an in-zone short circuit, it is significantly greater.

[0131] Set value DPM set Depends on the number of segments of the current sequence. If the amount of data is N, the number of segments is This is a common empirical value. This ensures that each segment of data is large enough to extract effective trend information and avoids the symbol sequence from being simplistic. According to the current sequence data with a data window of 10ms extracted from the actual system, N = 500, so the number of segments is selected as 100, and the setting value DPM is set =30, it is considered that when the dynamic mode matching distance is greater than 30, the phase current is faulty.

[0132] F2: Identification of internal and external faults includes:

[0133] Calculate the dynamic pattern matching distance of the three-phase current respectively; compare the dynamic pattern matching distance of each phase with the setting value 30; identify the specific fault phase and output the corresponding protection action.

[0134] Executes the criteria for identifying internal and external faults. The system calculates the dynamic pattern matching distances for the A, B, and C phases and compares each calculated result with a set value of 30. When the dynamic pattern matching distance for a phase is greater than the set value, it is determined that an internal fault has occurred on that phase. When the dynamic pattern matching distances for all phases are less than the set value, it is determined to be normal operation or an external fault state.

[0135] The criteria for identifying faults inside and outside the zone are:

[0136] like It is determined that a short circuit fault has occurred in the area and a trip command is sent;

[0137] Otherwise, it is judged to be normal operation or an out-of-zone short circuit fault, and the protection device is reset.

[0138] F3: Specific fault phase distinction includes the ability to identify single-phase grounding faults, two-phase short circuit faults, two-phase grounding short circuit faults, and three-phase short circuit faults, and maintain identification accuracy under different signal-to-noise ratios and high transition resistance conditions.

[0139] Identify the specific fault type and output protection action. According to the calculation results of the dynamic mode matching distance of each phase, the system can accurately identify single-phase grounding faults (such as when phase A is grounded, DPM A >30 DPM B , DPM C <30), two-phase short circuit fault (such as AB phase short circuit, DPM A , DPM B >30 DPM C <30), two-phase ground short circuit fault and three-phase short circuit fault and other different fault types. Figure 3The fault current waveforms shown in the figure show that when a three-phase short circuit occurs at point F1, the current waveforms on the PV and system sides differ significantly, causing the dynamic mode matching distances for each phase to be significantly greater than the set value. Once an internal fault is confirmed, the system immediately issues a trip command to disconnect the faulty line.

[0140] In an optional implementation, the fault judgment in step S600 may also be combined with multiple criteria, such as a traditional current differential criterion or an impedance criterion, to form a comprehensive protection solution to further improve the reliability of protection.

[0141] In another optional implementation, the system may also select different protection action strategies according to the severity and type of the fault. For example, for a minor fault, an alarm signal may be issued first, while for a serious fault, a tripping operation may be performed immediately.

[0142] It should be noted that the fault identification method of the present invention has excellent anti-interference capabilities. As shown in Table 2, under different signal-to-noise ratio conditions (5dB, 10dB, and 20dB), for each type of fault (AG, AB, ABG, and ABC), the DPM value of the fault phase is significantly greater than the set value of 30, enabling accurate identification of the fault phase and demonstrating significant noise interference resistance. Furthermore, as shown in Table 3, under different transition resistance conditions (50Ω, 100Ω, and 200Ω), this method can correctly identify both in-zone and out-of-zone faults and the specific fault phase, demonstrating excellent resistance to transition resistance.

[0143] In summary, through the complete technical solution of S100-S600 mentioned above, the present invention realizes reliable protection of new energy transmission lines, effectively solves the technical difficulties faced by traditional protection methods after the access of new energy stations, and improves the sensitivity, reliability and speed of protection.

[0144] Example 3

[0145] Through the electromagnetic transient simulation software PSCAD to build Figure 3 The photovoltaic AC line simulation model shown is used to simulate and verify the new energy transmission line protection method based on dynamic mode matching proposed in this embodiment:

[0146] 1) Build a model

[0147] The simulation model of 220kV photovoltaic AC line is as follows: Figure 3 As shown in the figure, the system frequency is 50Hz, the sampling frequency is 50kHz, and the data window length is 10ms. The GFM uses a virtual synchronous machine and virtual impedance current limiting control strategy. The PV station uses a single-machine multiplier model with a capacity of 100MVA and an AC line length of 40km. Three fault points are set in the simulation model: F1 (internal fault), F2 (external fault on the PV side), and F3 (external fault on the grid side).

[0148] 2) Simulation analysis

[0149] a) Protection principle verification

[0150] In order to verify the effectiveness of the protection method for renewable energy transmission lines based on dynamic pattern matching proposed in this embodiment, Figure 3 A three-phase short circuit fault is set at point F1, and the fault current waveform is as follows: Figure 5 shown.

[0151] Among them I wφ and I sφ The φ-phase sampled currents are for the PV and system sides, respectively. Since the positive direction of the AC line current transformers is from the busbar to the line, during normal operation, the phase currents on both sides of the AC line are symmetrical about the zero-value horizontal axis. At this point, flipping the current waveform on one side about the zero-value horizontal axis makes the currents on both sides almost coincide, with extremely similar current shapes. However, after a short-circuit fault (0.5s) occurs in the AC line, the frequency, amplitude, and phase of the fault current on the PV side clearly change significantly, differing significantly from the waveform on the AC grid side.

[0152] exist Figure 1 At point F1, single-phase ground fault (AG), two-phase short-circuit fault (AB), two-phase ground short-circuit fault (ABG), and three-phase short-circuit fault (ABC) were set, and the transition resistance was set to 5Ω. To verify the noise immunity of the protection method, Gaussian white noise was introduced into the current sampling data, with signal-to-noise ratios of 5, 10, and 20 dB, respectively. The simulation results are shown in Table 2.

[0153] Table 2 Fault identification results under different signal-to-noise ratios

[0154]

[0155] From Table 2, we can see that for all kinds of faults occurring in the area, the fault phase Both are significantly greater than the set value of 30. The protection can accurately and sensitively identify the fault phase and has significant anti-noise interference capabilities.

[0156] b) Performance of the proposed protection method for high transition resistance ground faults

[0157] To further verify the protection method's performance against transition resistance during both intra- and extra-zone faults, a single-phase ground fault on phase A was assumed. Considering a 220 kV transmission line with a maximum transition resistance of 100 Ω, a single-phase ground fault on phase A was set for F1, F2, and F3, with transition resistances of 50, 100, and 200 Ω, respectively. The simulation results are shown in Table 3.

[0158] Table 3 Fault identification results under different transition resistances

[0159]

[0160] As shown in Table 3, under the influence of different transition resistances, the proposed protection method can correctly identify the faults inside and outside the zone and the specific fault phase, and has good ability to resist transition resistance.

[0161] Example 4

[0162] The above is a schematic scheme of a dynamic pattern recognition and protection method for a new energy transmission line. It should be noted that the technical scheme of the dynamic pattern recognition and protection system for a new energy transmission line and the technical scheme of the dynamic pattern recognition and protection method for a new energy transmission line are based on the same concept. For details not described in detail in the technical scheme of the dynamic pattern recognition and protection system for a new energy transmission line in this embodiment, please refer to the description of the technical scheme of the dynamic pattern recognition and protection method for a new energy transmission line.

[0163] This embodiment further provides a new energy transmission line dynamic pattern recognition protection system, including:

[0164] The protection algorithm startup unit is used to obtain the three-phase current at the protection installation in real time, calculate the phase current mutation amount, and start the protection algorithm when the mutation amount exceeds the threshold;

[0165] A communication unit, used for sending current sampling data to a protection device at the opposite end;

[0166] A dynamic pattern matching distance calculation unit is used to obtain current sampling data of the opposite side protection device and calculate the dynamic pattern matching distance;

[0167] The fault judgment unit is used to compare the size relationship between the action amount and the set value, judge the fault inside and outside the zone and output the protection action instruction.

[0168] This embodiment also provides an electronic device suitable for dynamic pattern recognition protection of new energy transmission lines, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the dynamic pattern recognition protection method for new energy transmission lines proposed in the above embodiment.

[0169] This embodiment further provides a storage medium storing a computer program, which, when executed by a processor, implements the method for dynamic pattern recognition and protection of new energy transmission lines proposed in the above embodiment.

[0170] The storage medium proposed in this embodiment and the method for implementing dynamic pattern recognition and protection of new energy transmission lines proposed in the above embodiment belong to the same inventive concept. Technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0171] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general hardware, and of course can also be implemented by hardware. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a computer's floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0172] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A new energy transmission line dynamic pattern recognition protection method, characterized by: include, Determine whether to start the protection algorithm based on the phase current mutation amount; Obtain the three-phase current sampling values ​​at the protection installation location and send the local current sampling data to the opposite protection device; After obtaining the current sampling data of the opposite end protection device, the current sampling sequence data is divided into segment modes; The mean information of the segmented current sequence is combined with the change direction to generate an improved pattern symbol based on the ternary pattern; Calculate the dynamic mode matching distance between the currents at both ends; The dynamic mode matching distance is compared with the preset setting value. If it is greater than the setting value, it is judged that an internal short circuit fault has occurred and a trip command is sent. Otherwise, it is judged to be normal operation or an external short circuit fault.

2. The method for dynamic pattern recognition and protection of a new energy transmission line according to claim 1, characterized in that: The determining whether to start the protection algorithm based on the phase current mutation amount includes: Calculate the phase current mutation in real time; start the protection algorithm when the mutation of any phase current exceeds 0.1 times the load current; The protection starts when three consecutive sampling points meet the start criteria, otherwise the protection is reset.

3. The method for dynamic pattern recognition and protection of a new energy power transmission line according to claim 2, characterized in that: The generating of the improved pattern symbol based on the ternary pattern comprises: The pattern symbol type is determined according to the mean value and slope positivity of the current sequence; The three states of rising, falling and holding correspond to different mode symbols respectively; The symbols of the hold states under different mean values ​​are unified as the same identifier to improve computational efficiency.

4. The method for dynamic pattern recognition and protection of a new energy power transmission line according to claim 3, characterized in that: The calculating of the dynamic mode matching distance between the currents at both ends includes: After converting the current sampling data into pattern symbols, the pattern matching distance is calculated; a morphological pattern matrix between two sets of morphological pattern sequences is constructed; A dynamic programming algorithm is used to calculate the cumulative distance matrix; the dynamic pattern matching distance is determined by the final value of the cumulative distance matrix.

5. The method for dynamic pattern recognition and protection of new energy transmission lines according to claim 4, characterized in that: The preset setting value is determined according to the number of segments of the current sequence: According to the current sequence data extracted from the 10ms data window, the number of segments is determined to be 100; the dynamic mode matching setting value is set to 30; When the dynamic mode matching distance is greater than 30, it is determined that the phase current has a fault.

6. The method for dynamic pattern recognition and protection of new energy transmission lines according to claim 5, characterized in that: The in-zone and out-zone fault identification includes: Calculate the dynamic mode matching distance of the three-phase current respectively; compare the dynamic mode matching distance of each phase with the setting value 30; identify the specific fault phase and output the corresponding protection action.

7. The method for dynamic pattern recognition and protection of a new energy power transmission line according to claim 6, characterized in that: The specific fault phase distinction includes the ability to identify single-phase grounding faults, two-phase short circuit faults, two-phase grounding short circuit faults and three-phase short circuit faults, and maintain identification accuracy under different signal-to-noise ratios and high transition resistance conditions.

8. A dynamic pattern recognition and protection system for a new energy transmission line, based on the dynamic pattern recognition and protection method for a new energy transmission line according to any one of claims 1 to 7, characterized in that: The protection algorithm starting unit is further included, which is used to obtain the three-phase current at the protection installation in real time, calculate the phase current mutation amount, and start the protection algorithm when the mutation amount exceeds a threshold; A communication unit, used for sending current sampling data to a protection device at the opposite end; A dynamic pattern matching distance calculation unit is used to obtain current sampling data of the opposite side protection device and calculate the dynamic pattern matching distance; The fault judgment unit is used to compare the size relationship between the action amount and the set value, judge the fault inside and outside the zone and output the protection action instruction.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the new energy transmission line dynamic pattern recognition protection method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the dynamic pattern recognition protection method for new energy transmission lines according to any one of claims 1 to 7 are implemented.

Images