A method, device, program product and medium for self-setting of a relay protection parameter

Abstract

A relay protection parameter self-configuration method, device, program product and medium, relate to the technical field of power system. In the method, three-phase voltage, three-phase current and neutral point voltage of a target generator in a preset time period are obtained; a phase angle difference is determined based on the three-phase voltage and the three-phase current, and a comprehensive state feature vector is constructed based on the phase angle difference and the neutral point voltage; the operating conditions of the target generator are classified to obtain multiple operating modes, and a corresponding health state trajectory cluster is determined according to the comprehensive state feature vector in each operating mode; the geometric characteristics of each health state trajectory cluster are analyzed, and the protection parameters of a relay protection device for protecting the target generator are determined based on each geometric characteristic. The method has the effect of improving the effectiveness of self-configuration protection parameters of the relay protection device.

Description

Technical Field

[0001] This application relates to the field of power system technology, specifically to a method, equipment, program product, and medium for self-matching relay protection parameters. Background Technology

[0002] As the requirements for safe and stable operation of power systems become increasingly stringent, protection technologies for large generator units are developing towards intelligence and self-adaptation. Automatic configuration and optimization of relay protection parameters, i.e., parameter self-configuration, is a key technology for achieving intelligent and self-adaptive generator protection. It aims to replace cumbersome manual settings, improve the response speed and accuracy of protection, and cope with the increasingly complex operating environment of the power grid.

[0003] Currently, mainstream methods for automatically configuring relay protection parameters typically rely on a pre-defined power grid topology and simplified generator fault models. These methods determine protection settings through large-scale simulated short-circuit calculations. For example, by simulating different types of short-circuit faults, the corresponding changes in electrical quantities are calculated, and the operating thresholds and delays for protections such as negative-sequence overvoltage or overcurrent protection are automatically generated accordingly.

[0004] However, the aforementioned model-based self-matching methods, due to their simplification of generator fault characteristics and neglect of the complexity of actual operating conditions, struggle to effectively model extremely subtle faults such as initial inter-turn short circuits within the generator. This results in inherent defects in the automatically generated protection settings: if the settings are set too low to ensure sensitivity, they are highly susceptible to malfunctions caused by measurement noise during normal operation or external power grid disturbances; if the settings are set too high to avoid disturbances, they will fail to protect against genuine early-stage, subtle faults. Therefore, existing self-matching methods for protection parameters are ineffective in addressing such faults. Summary of the Invention

[0005] This application provides a method, equipment, program product, and medium for self-matching relay protection parameters, which improves the effectiveness of self-matching protection parameters in relay protection devices.

[0006] The first aspect of this application provides a method for self-matching relay protection parameters, specifically including: Obtain the three-phase voltage, three-phase current, and neutral point voltage of the target generator within a preset time period; The phase angle difference is determined based on the three-phase voltage and three-phase current, and a comprehensive state feature vector is constructed based on the phase angle difference and the neutral point voltage. The operating conditions of the target generator are classified to obtain multiple operating modes. Based on the comprehensive state feature vector of each operating mode, the corresponding health state trajectory cluster is determined. The geometric characteristics of each health state trajectory cluster are analyzed, and the protection parameters of the relay protection device for protecting the target generator are determined based on each geometric characteristic.

[0007] By adopting the above technical solution, the three-phase voltage, three-phase current, and neutral point voltage of the target generator within a preset time period are obtained. These key electrical parameters provide a guarantee for the comprehensive collection of basic data information on the generator's operating status. The phase angle difference is determined based on the three-phase voltage and three-phase current, and the phase angle difference is combined with the neutral point voltage to construct a comprehensive state feature vector. The multi-dimensional information contained in the feature vector enables the generator's operating status to be accurately quantified and characterized. The operating conditions of the target generator are classified into multiple operating modes, and the corresponding health status trajectory clusters are determined according to the comprehensive state feature vectors under each operating mode. The correspondence between operating modes and health statuses constructs a health status assessment benchmark for the generator under different operating modes. The geometric characteristics of each health status trajectory cluster are analyzed, and the protection parameters of the relay protection device used to protect the target generator are determined based on each geometric characteristic. The protection parameters fully integrate the health status characteristics of the generator under different operating modes, thereby improving the effectiveness of the self-matching protection parameters of the relay protection device.

[0008] Optionally, the step of determining the phase angle difference based on the three-phase voltage and three-phase current, and constructing a comprehensive state feature vector based on the phase angle difference and the neutral point voltage, specifically includes: Dynamic window sliding sampling is performed on the three-phase voltage and the three-phase current to obtain the voltage and current sampling sequence within a preset sampling period; The voltage and current sampling sequence is subjected to wavelet transform to obtain a time-frequency distribution feature map, and the fundamental frequency component and harmonic component of the three-phase voltage and three-phase current are extracted from the time-frequency distribution feature map respectively. Based on each of the fundamental frequency components, the phase angle difference between the three-phase voltage and the three-phase current is determined; A comprehensive state feature vector is constructed based on the phase angle difference, the amplitude of the harmonic components, and the neutral point voltage.

[0009] By adopting the above technical solution, dynamic window sliding sampling is performed on the three-phase voltage and three-phase current to obtain the voltage and current sampling sequence within a preset sampling period, realizing continuous tracking and acquisition of electrical signals. Wavelet transform is performed on the voltage and current sampling sequence to obtain a time-frequency distribution feature map, which fully presents the characteristics of the voltage and current signals in the time and frequency domains. Then, the fundamental frequency components and harmonic components of the three-phase voltage and three-phase current are extracted from the time-frequency distribution feature map to obtain the frequency composition of the electrical signals. Based on each fundamental frequency component, the phase angle difference between the three-phase voltage and three-phase current is determined, which accurately reflects the phase relationship between voltage and current. The phase angle difference, the amplitude of the harmonic components, and the neutral point voltage are constructed into a comprehensive state feature vector, which provides a reliable data foundation for subsequent analysis of various fault characteristics of the generator and improves the accuracy of fault feature identification.

[0010] Optionally, determining the phase angle difference between the three-phase voltage and the three-phase current based on each of the fundamental frequency components specifically includes: Perform a Hilbert transform on each of the fundamental frequency components to obtain the corresponding orthogonal components; Based on each fundamental frequency component and its corresponding quadrature component, the phase angles of the three-phase voltage and three-phase current are determined respectively. Based on each of the aforementioned phase angles, the initial phase difference between the three-phase voltage and the three-phase current is determined; The initial phase difference is subjected to a moving average filter to obtain the phase angle difference between the three-phase voltage and the corresponding three-phase current.

[0011] By adopting the above technical solution, Hilbert transform is performed on each fundamental frequency component to obtain the corresponding orthogonal components, providing a mathematical basis for calculating the phase angle of voltage and current signals. Based on each fundamental frequency component and the corresponding orthogonal components, the phase angles of the three-phase voltage and three-phase current are determined respectively, realizing accurate calculation of the instantaneous phase of voltage and current. Based on each phase angle, the initial phase difference between the three-phase voltage and the three-phase current is determined, obtaining the original phase difference information. The initial phase difference is subjected to moving average filtering to obtain the phase angle difference between the three-phase voltage and the corresponding three-phase current, eliminating noise interference in the phase difference signal and improving the stability and reliability of the phase angle difference calculation results.

[0012] Optionally, the construction of a comprehensive state feature vector based on the phase angle difference, the amplitude of the harmonic components, and the neutral point voltage specifically includes: Based on the variation range of the phase angle difference within a preset time period, a phase angle stability index is determined; Based on the consistency between the amplitude of the harmonic components and the changing trend of the neutral point voltage, the fault sensitivity index is determined. The phase angle stability index and the fault sensitivity index are combined to obtain a comprehensive state feature vector.

[0013] By adopting the above technical solution, the phase angle stability index is determined based on the change range of the phase angle difference within a preset time, which quantifies the stability of the generator's operating state; the fault sensitivity index is determined based on the consistency between the amplitude of the harmonic components and the change trend of the neutral point voltage, which reflects the response characteristics of electrical parameters to fault conditions; the phase angle stability index and the fault sensitivity index are combined to obtain a comprehensive state feature vector, which contains both the stability and fault sensitivity information of the generator operation, thus enhancing the ability to identify abnormal generator operating states.

[0014] Optionally, the step of performing wavelet transform on the voltage and current sampling sequence to obtain a time-frequency distribution feature map, and extracting the fundamental frequency component and harmonic component of the three-phase voltage and three-phase current from the time-frequency distribution feature map, specifically includes: performing wavelet transform hierarchical decomposition on the voltage and current sampling sequence of the three-phase voltage and three-phase current to obtain wavelet coefficients of different scales; Based on the wavelet coefficients, a time-frequency distribution feature map reflecting the change of the frequency components of the three-phase voltage and three-phase current over time is obtained; In the time-frequency distribution feature map, the energy distribution ratio of each frequency band is calculated, and the frequency band with the largest energy ratio is determined as the fundamental frequency component. In the time-frequency distribution feature map, other frequency bands outside the frequency band where the fundamental frequency component is located are identified as harmonic components.

[0015] By employing the above technical solution, wavelet transform is used to decompose the voltage and current sampling sequences of three-phase voltage and three-phase current into wavelet coefficients at different scales, achieving fine decomposition of electrical signals in different frequency bands. Based on the wavelet coefficients, a time-frequency distribution feature map reflecting the frequency components of three-phase voltage and three-phase current changing over time is obtained, showing the complete change process of electrical signals in the time and frequency domains. The energy distribution ratio of each frequency band is calculated in the time-frequency distribution feature map, and the frequency band with the largest energy ratio is determined as the fundamental frequency component, realizing the accurate extraction of the main frequency components in the electrical signal. Other frequency bands outside the fundamental frequency component are determined as harmonic components, effectively separating the fundamental frequency component and harmonic components, and improving the accuracy of subsequent phase angle difference and eigenvector calculations.

[0016] Optionally, classifying the operating conditions of the target generator to obtain multiple operating modes, and determining the corresponding health state trajectory cluster based on the comprehensive state feature vector of each operating mode, specifically includes: The operating conditions of the target generator are classified based on active power and reactive power to obtain multiple operating modes; Within the target time period, record the continuous trajectory of the comprehensive state feature vector changing over time in each of the aforementioned operating modes; Based on the continuous trajectories of each operating mode, a cluster of health status trajectories within each operating mode is determined.

[0017] By adopting the above technical solution, the operating conditions of the target generator are classified based on active power and reactive power, resulting in multiple operating modes and establishing classification standards for different operating states of the generator. Within the target time period, the continuous trajectory of the comprehensive state feature vector within each operating mode is recorded as a function of time, reflecting the dynamic change law of the generator's operating characteristics under different power levels. Based on the continuous trajectory of each operating mode, the health state trajectory cluster within each operating mode is determined, and the health state boundary of the generator under different power levels is constructed, improving the adaptability and accuracy of health state assessment.

[0018] Optionally, the step of analyzing the geometric features of each of the health state trajectory clusters and determining the corresponding protection parameters of the relay protection device for protecting the target generator based on each of the geometric features specifically includes: Calculate the geometric center coordinates and boundary curves of each health state trajectory cluster; Statistically analyze the distribution density and variance of trajectory points within each health state trajectory cluster; The starting current threshold and starting voltage threshold of the relay protection device used to protect the target generator are determined based on the magnitude and distribution of the geometric center coordinates. The range of the protection characteristic curve of the relay protection device is determined based on the range of the boundary curve, and the range of the protection characteristic curve characterizes the safe operating range of the target generator's operating parameters; The tripping delay time of the relay protection device is determined based on the distribution density and variance. The starting current threshold, starting voltage threshold, protection characteristic curve range, and trip delay time of the relay protection device are used as protection parameters for the relay protection device used to protect the target generator.

[0019] By adopting the above technical solution, the geometric center coordinates and boundary curves of each health state trajectory cluster are calculated, and the mathematical characteristics of the generator's healthy operating state are obtained. The distribution density and variance of trajectory points within each health state trajectory cluster are statistically analyzed, quantifying the fluctuation pattern of the generator's operating state. The starting current threshold and starting voltage threshold of the relay protection device are determined based on the magnitude and distribution position of the geometric center coordinates, so that the protection starting value is adapted to the actual operating characteristics of the generator. The range of the protection characteristic curve of the relay protection device is determined based on the range of the boundary curve, realizing the precise definition of the safe range of generator operating parameters. The tripping delay time of the relay protection device is determined based on the distribution density and variance, avoiding maloperation caused by fluctuations in the generator's operating state. The starting current threshold, starting voltage threshold, protection characteristic curve range, and tripping delay time of the relay protection device are used as protection parameters, realizing the comprehensive adaptive configuration of relay protection parameters.

[0020] In a second aspect, this application provides an electronic device for self-matching relay protection parameters, the device comprising: one or more processors and a memory; the memory being coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors calling the computer instructions to cause the electronic device for self-matching relay protection parameters to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, this application provides a computer program product containing instructions that, when the computer program product is run on an electronic device with self-matching relay protection parameters, cause the electronic device to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, this application provides a computer-readable storage medium including instructions that, when executed on an electronic device with self-matched relay protection parameters, cause the device to perform the method described in the first aspect and any possible implementation thereof. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the architecture of a relay protection parameter self-matching system provided in an embodiment of this application; Figure 2 This is a flowchart illustrating a relay protection parameter self-matching method provided in an embodiment of this application; Figure 3 This is a schematic diagram of signal time-frequency feature analysis provided in an embodiment of this application; Figure 4 This is a schematic diagram illustrating the construction and analysis process of the health status trajectory cluster provided in the embodiments of this application; Figure 5This is an exemplary hardware structure diagram of an electronic device for self-matching relay protection parameters provided in an embodiment of this application. Detailed Implementation

[0024] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0025] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0026] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0027] Figure 1 An exemplary system architecture for a relay protection parameter self-matching system is shown.

[0028] like Figure 1 As shown, the system architecture may include electronic device 11, network 12, and relay protection device 13. Network 12 serves as the medium for providing a communication link between electronic device 11 and relay protection device 13. Network 12 may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0029] The relay protection device 13 can transmit the collected three-phase voltage, three-phase current, and neutral point voltage data to the electronic device 11 via the network 12. The electronic device 11 is equipped with a protection parameter self-configuration application, which can perform feature extraction, operating condition classification, and parameter calculation on the received data.

[0030] Electronic device 11 is hardware and can be any electronic device with computing capabilities, including but not limited to engineering laptops, workstations, servers, and computing centers. Electronic device 11 can analyze and process the collected data and generate protection parameter configuration instructions.

[0031] The relay protection device 13 can be a digital protection device used to collect generator operating data and perform protection functions. The relay protection device 13 receives protection parameter configuration instructions from the electronic device 11 and updates local protection parameters according to the configuration instructions to realize the protection function for the generator.

[0032] The following detailed explanation uses the electronic device side as an example.

[0033] This embodiment provides a method for self-configuration of relay protection parameters. Figure 2 This is a flowchart illustrating a relay protection parameter self-matching method provided in an embodiment of this application, as shown below. Figure 2 As shown, the method includes S101 to S104: S101: Obtain the three-phase voltage, three-phase current and neutral point voltage of the target generator within a preset time period.

[0034] In this embodiment of the application, the target generator refers to a power generation device that requires self-configuration of relay protection parameters; the three-phase voltage represents the voltage at the output terminals of the three physically distributed phase windings of the target generator, with the three phases distributed sequentially according to a pre-set phase order; the three-phase current represents the current signal, phase order, and voltage signal at the output terminals of the three phase windings of the target generator; and the neutral point voltage represents the voltage signal of the generator neutral point to ground.

[0035] Specifically, the electronic equipment determines a preset time period by pre-setting the start and end times of data acquisition, and receives the three-phase voltage and three-phase current sample values ​​of the target generator collected by the relay protection device. The relay protection device collects the neutral point voltage value of the target generator to ground as the neutral point voltage, and performs a transformation ratio conversion on the three-phase voltage and three-phase current sample values ​​through voltage transformers and current transformers. The converted three-phase voltage, three-phase current, and neutral point voltage values ​​are then converted into digital signals and transmitted to the electronic equipment. The electronic equipment stores the received three-phase voltage, three-phase current, and neutral point voltage according to a preset sampling frequency.

[0036] S102: Determine the phase angle difference based on the three-phase voltage and three-phase current, and construct a comprehensive state feature vector based on the phase angle difference and neutral point voltage.

[0037] In this embodiment, the phase angle difference represents the phase angle difference between the three-phase voltage and the three-phase current. The phase angle difference is used to reflect the timing relationship between voltage and current under the generator operating state. The comprehensive state feature vector represents a multi-dimensional data combination containing phase angle difference information and neutral point voltage information. The comprehensive state feature vector is used to characterize the generator operating state characteristics.

[0038] Specifically, the electronic equipment performs dynamic window sliding sampling on the stored three-phase voltage and three-phase current. The sampling window length is set to one power frequency cycle. The sampled voltage and current sequences are sent to the wavelet transform module for hierarchical decomposition to obtain a time-frequency distribution feature map reflecting the change of voltage and current frequency components over time. The fundamental frequency components and harmonic components of the three-phase voltage and three-phase current are extracted from the time-frequency distribution feature map. The fundamental frequency components are subjected to Hilbert transform to obtain orthogonal components. The phase angles of the three-phase voltage and three-phase current are calculated using the fundamental frequency components and orthogonal components. The initial phase difference between the three-phase voltage and the corresponding three-phase current is calculated. The initial phase difference is subjected to moving average filtering to obtain the phase angle difference. The phase angle difference is combined with the neutral point voltage to construct a comprehensive state feature vector.

[0039] Based on the above embodiments, as an optional embodiment, the phase angle difference is determined based on the three-phase voltage and three-phase current, and a comprehensive state feature vector is constructed based on the phase angle difference and the neutral point voltage. This step may further include steps S201 to S204: S201: Perform dynamic window sliding sampling on the three-phase voltage and three-phase current to obtain the voltage and current sampling sequence within the preset sampling period.

[0040] Specifically, the electronic device sets the dynamic sampling window length to one power frequency cycle, slides the sampling window continuously on the time axis, and performs data acquisition once each time the sampling window slides to a new position, collecting three-phase voltage sampling values ​​and three-phase current sampling values. The electronic device receives the sampling data transmitted by the relay protection device and arranges the sampling data in the order of sampling time to form a voltage and current sampling sequence. The sampling sequence includes a three-phase voltage value sequence and a three-phase current value sequence.

[0041] For example, the electronic device continuously moves the dynamic sampling window on the time axis, collecting data once every power frequency cycle. It collects the voltage and current sampling values ​​of phase A, the voltage and current sampling values ​​of phase B, the voltage and current sampling values ​​of phase C, and arranges the collected voltage and current sampling values ​​in chronological order to form the voltage and current sampling sequences of phase A, phase B, and phase C.

[0042] S202: Perform wavelet transform on the voltage and current sampling sequences to obtain the time-frequency distribution feature map, and extract the fundamental frequency components and harmonic components of the three-phase voltage and three-phase current from the time-frequency distribution feature map respectively.

[0043] In this embodiment, wavelet transform represents a mathematical transformation method for decomposing voltage and current sampling sequences into components of different frequency scales. Wavelet transform is used to analyze the time-frequency characteristics of voltage and current signals. The time-frequency distribution feature map represents a two-dimensional spectrum reflecting the change of frequency components of voltage and current signals over time. The fundamental frequency component represents the power frequency component in the voltage and current signal. The harmonic components represent the frequency components in the voltage and current signal that are higher than the power frequency.

[0044] Specifically, the electronic equipment uses wavelet basis functions to perform hierarchical decomposition on the voltage and current sampling sequences, obtaining wavelet coefficients at different scales. Through wavelet coefficient reconstruction, a time-frequency distribution feature map reflecting the frequency components of three-phase voltage and three-phase current changing with time is obtained. The energy distribution ratio of each frequency band in the time-frequency distribution feature map is analyzed, and the frequency band with the largest energy ratio is determined as the frequency band where the fundamental frequency component is located. The frequency bands outside the frequency band where the fundamental frequency component is located are determined as the frequency band where the harmonic components are located. The fundamental frequency component and harmonic component of three-phase voltage and three-phase current are extracted from each frequency band.

[0045] Figure 3 This is a schematic diagram of signal time-frequency characteristic analysis provided in an embodiment of this application. The diagram illustrates, in a visual manner, the technical concept of decomposing the original time-domain signal (left) into different frequency components (right) through core time-frequency analysis methods (arrows in the diagram).

[0046] Figure 3 The chart on the left shows the raw signal waveform acquired in the time domain, with amplitude on the vertical axis and time on the horizontal axis. This waveform consists of a smooth fundamental signal and a fault disturbance signal superimposed on it, representing the voltage or current sampling sequence to be analyzed.

[0047] The arrow in the middle of the figure indicates wavelet transform, signifying that wavelet transform processing is performed on the sampled sequence on the left. This step corresponds to the core step in the method flow of this application, namely, mapping the one-dimensional time-domain signal to a two-dimensional time-frequency plane through wavelet transform, thereby obtaining a time-frequency distribution feature map that reflects the change of signal frequency components over time.

[0048] Figure 3 The chart on the right is a schematic representation of the different frequency components extracted after analyzing this time-frequency distribution feature map. It shows that by analyzing this feature map, the original composite signal can be clearly decomposed into the following three components: Fundamental frequency component: This corresponds to the frequency band with the largest energy proportion in the time-frequency distribution characteristic diagram, which exists continuously across the entire time axis. This component represents the power frequency component in the signal.

[0049] Steady-state harmonic components: These are the other frequency bands that exist continuously along the entire time axis, outside the frequency band of the fundamental frequency component, as shown in the time-frequency distribution characteristic diagram. These are the inherent harmonics during normal system operation.

[0050] Transient harmonic components: These are harmonic frequency bands that appear only during specific time periods (consistent with the occurrence time of the "fault disturbance" in the left figure) or whose energy is significantly enhanced, corresponding to the time-frequency distribution characteristic map. This component is a key feature for fault identification.

[0051] In summary, Figure 3 The technical solution of this application is clearly illustrated: First, a time-frequency distribution feature map is generated through wavelet transform, achieving a complete representation of the signal in both the time and frequency domains. Then, by analyzing this feature map, the fault-related "transient harmonic components" can be accurately separated from the "fundamental frequency components" and "steady-state harmonic components." This process provides high-quality and highly discriminative data input for the subsequent construction of a comprehensive state feature vector and the realization of accurate relay protection parameter self-matching.

[0052] Based on the above embodiments, as an optional embodiment, wavelet transform is performed on the voltage and current sampling sequences to obtain a time-frequency distribution feature map, and the fundamental frequency components and harmonic components of the three-phase voltage and three-phase current are extracted from the time-frequency distribution feature map respectively. This step may also include steps S301 to S304: S301: Perform wavelet transform hierarchical decomposition on the voltage and current sampling sequences of three-phase voltage and three-phase current to obtain wavelet coefficients at different scales.

[0053] Specifically, the electronic device uses wavelet basis functions as decomposition tools to perform high-pass and low-pass filtering operations on the voltage and current sampling sequences. The high-pass filter output is used as the first layer of high-frequency components, and the low-pass filter output is used as the first layer of low-frequency components. High-pass and low-pass filtering operations are then performed on the first layer of low-frequency components to obtain the second layer of high-frequency and low-frequency components. The high-pass and low-pass filtering operations are repeated until the preset number of decomposition layers is completed. The high-frequency and low-frequency components obtained from each decomposition layer are used as wavelet coefficients of different scales.

[0054] For example, the electronic device uses wavelet basis functions to perform wavelet decomposition on the A-phase voltage sampling sequence. The first high-pass filtering and low-pass filtering operations are performed to obtain the first-level high-frequency wavelet coefficients and low-frequency wavelet coefficients. The second high-pass filtering and low-pass filtering operations are performed on the first-level low-frequency wavelet coefficients to obtain the second-level high-frequency wavelet coefficients and low-frequency wavelet coefficients. The high-pass filtering and low-pass filtering operations are repeated to complete the preset number of decomposition levels. The same decomposition process is performed on the B-phase voltage sampling sequence to obtain wavelet coefficients of different scales. The same decomposition process is performed on the C-phase voltage sampling sequence to obtain wavelet coefficients of different scales. The same decomposition process is performed on the three-phase current sampling sequence to obtain wavelet coefficients of different scales.

[0055] S302: Based on wavelet coefficient reconstruction, a time-frequency distribution characteristic map reflecting the change of frequency components of three-phase voltage and three-phase current over time is obtained.

[0056] Specifically, the electronic device retains the wavelet coefficients of the corresponding layer of the frequency band to be analyzed, sets the wavelet coefficients of layers outside the frequency band to be analyzed to zero, performs inverse wavelet transform to reconstruct the signal components of the frequency band to be analyzed, repeats the reconstruction process until the signal components of all frequency bands are reconstructed, and arranges the reconstructed signal components of each frequency band on the time axis and frequency axis to generate a time-frequency distribution characteristic map that reflects the change of the frequency components of the three-phase voltage and three-phase current over time.

[0057] For example, the electronic device retains the first-level high-frequency wavelet coefficients of phase A voltage, sets the wavelet coefficients from the second to the last level to zero, and performs inverse wavelet transform reconstruction to obtain the first frequency band signal component of phase A voltage. It retains the second-level high-frequency wavelet coefficients of phase A voltage, sets the wavelet coefficients from the first to the last level to zero, and performs inverse wavelet transform reconstruction to obtain the second frequency band signal component of phase A voltage. The reconstruction process is repeated to complete the reconstruction of all frequency band signal components. The reconstructed signal components are arranged on the time axis and frequency axis to obtain the time-frequency distribution feature map of phase A voltage. The same reconstruction process is performed on the wavelet coefficients of phase B voltage to obtain the time-frequency distribution feature map of phase B voltage. The same reconstruction process is performed on the wavelet coefficients of phase C voltage to obtain the time-frequency distribution feature map of phase C voltage. The same reconstruction process is performed on the wavelet coefficients of the three-phase current to obtain the time-frequency distribution feature map of the three-phase current. S303: In the time-frequency distribution feature map, the energy distribution ratio of each frequency band is calculated, and the frequency band with the largest energy ratio is determined as the fundamental frequency component.

[0058] Specifically, the electronic equipment extracts the signal components of each frequency band from the time-frequency distribution feature map, calculates the sum of the squares of the signal components of each frequency band to obtain the frequency band energy value, calculates the sum of the energy values ​​of all frequency bands to obtain the total signal energy, divides the energy value of each frequency band by the total signal energy to obtain the energy distribution ratio of each frequency band, compares the magnitude of the energy distribution ratio of each frequency band, and determines the frequency band with the largest energy distribution ratio as the frequency band where the power frequency fundamental frequency component is located, and extracts the signal component of this frequency band as the fundamental frequency component of the three-phase voltage and three-phase current.

[0059] For example, an electronic device extracts the first frequency band signal component from the time-frequency distribution feature map of phase A voltage to calculate the energy value of the first frequency band, extracts the second frequency band signal component to calculate the energy value of the second frequency band, extracts all frequency band signal components to calculate the frequency band energy value, divides the energy value of each frequency band by the total energy of the phase A voltage signal to obtain the energy distribution ratio, compares the size of the energy distribution ratio of each frequency band, and determines the frequency band with the largest energy distribution ratio as the fundamental frequency component frequency band of phase A voltage. The same analysis process is performed on the time-frequency distribution feature map of phase B voltage to determine the fundamental frequency component frequency band of phase B voltage, the same analysis process is performed on the time-frequency distribution feature map of phase C voltage to determine the fundamental frequency component frequency band of phase C voltage, and the same analysis process is performed on the time-frequency distribution feature map of three-phase current to determine the fundamental frequency component frequency band of three-phase current.

[0060] In the time-frequency distribution characteristic map, S304 identifies frequency bands outside the band containing the fundamental frequency component as harmonic components.

[0061] Specifically, the electronic device locates the frequency band where the fundamental frequency component is located from the time-frequency distribution feature map, marks all frequency bands outside the fundamental frequency component frequency band as harmonic component frequency bands, extracts the signal components in the marked frequency bands as harmonic components, and superimposes the signal components of all harmonic component frequency bands to obtain the complete harmonic component.

[0062] S203: Determine the phase angle difference between the three-phase voltage and the three-phase current based on each fundamental frequency component.

[0063] Specifically, the electronic equipment performs Hilbert transform on the fundamental frequency components of the three-phase voltage and the fundamental frequency components of the three-phase current to obtain orthogonal components. The phase angle of the three-phase voltage is calculated using the fundamental frequency components of the three-phase voltage and the orthogonal components. The phase angle of the three-phase current is calculated using the fundamental frequency components of the three-phase current and the orthogonal components. The initial phase difference between the phase angle of the three-phase voltage and the corresponding phase angle of the three-phase current is calculated. The initial phase difference is subjected to moving average filtering to eliminate phase jumps, resulting in a stable phase angle difference between the three-phase voltage and the three-phase current.

[0064] Based on the above embodiments, as an optional embodiment, the phase angle difference between the three-phase voltage and the three-phase current is determined based on each fundamental frequency component. This step may further include steps S401 to S404: S401: Perform Hilbert transform on each fundamental frequency component to obtain the corresponding orthogonal components.

[0065] In this embodiment, the Hilbert transform represents a mathematical transformation method for converting a real signal into an analytic signal. The Hilbert transform is used to obtain the quadrature components of the signal. The quadrature components represent signal components that are 90 degrees out of phase with the original signal, and are used to calculate the instantaneous phase of the signal.

[0066] Specifically, the electronic device processes the fundamental frequency component signal through a Hilbert transform filter. The filter performs phase delay and amplitude preservation operations on the fundamental frequency component signal to obtain an orthogonal component signal with a phase lag of 90 degrees. The fundamental frequency component signal and the orthogonal component signal are combined to form an analytical signal, which contains the amplitude and phase information of the fundamental frequency component signal.

[0067] S402: Determine the phase angles of the three-phase voltage and three-phase current based on each fundamental frequency component and its corresponding quadrature component.

[0068] Specifically, the electronic equipment calculates the arctangent of the fundamental frequency component and the quadrature component of the three-phase voltage to obtain the instantaneous phase angle of the three-phase voltage, and calculates the arctangent of the fundamental frequency component and the quadrature component of the three-phase current to obtain the instantaneous phase angle of the three-phase current. The electronic equipment selects a unified reference point to map the phase angles of the three-phase voltage and the three-phase current to the range of zero to 360 degrees.

[0069] S403: Determine the initial phase difference between the three-phase voltage and the three-phase current based on each phase angle.

[0070] For ease of understanding, this application document defines three-phase voltage as including phase A voltage, phase B voltage, and phase C voltage, and three-phase current as including phase A current, phase B current, and phase C current.

[0071] In this embodiment, the initial phase difference represents the difference between the three-phase voltage phase angle and the corresponding three-phase current phase angle. The initial phase difference is used to represent the relative phase relationship between the voltage and current signals. The relative phase relationship indicates the timing characteristic of the voltage phase leading or lagging the current phase.

[0072] Specifically, the electronic equipment calculates the difference between the phase angle of phase A voltage and the phase angle of phase A current to obtain the initial phase difference of phase A. It calculates the phase angle of phase B voltage and subtracts the phase angle of phase B current to obtain the initial phase difference of phase B. It calculates the phase angle of phase C voltage and subtracts the phase angle of phase C current to obtain the initial phase difference of phase C. The electronic equipment maps the calculated initial phase difference to a range of zero to 360 degrees as the relative phase relationship between the three-phase voltage and the three-phase current.

[0073] S404: Perform a moving average filter on the initial phase difference to obtain the phase angle difference between the three-phase voltage and the corresponding three-phase current.

[0074] Specifically, the electronic device sets the length of the moving average filter window, slides the filter window on the initial phase difference sequence, calculates the arithmetic mean of the initial phase difference values ​​in each sliding window, and uses the average value calculated in the sliding window as the filter output at the center point of the window. The filter window slides to cover the entire initial phase difference sequence, and the electronic device uses the filter output sequence as the phase angle difference between the three-phase voltage and the corresponding three-phase current.

[0075] S204: Construct a comprehensive state feature vector based on phase angle difference, harmonic component amplitude, and neutral point voltage.

[0076] Specifically, the electronic equipment calculates the change in phase angle difference within a preset time to obtain the phase angle stability index, and calculates the consistency between the harmonic component amplitude and the neutral point voltage change trend to obtain the fault sensitivity index. The electronic equipment combines the phase angle stability index and the fault sensitivity index to construct a comprehensive state feature vector, which contains generator steady-state operation characteristics and fault characteristic information.

[0077] Based on the above embodiments, as an optional embodiment, a comprehensive state feature vector is constructed based on the phase angle difference, the amplitude of harmonic components, and the neutral point voltage. This step may further include steps S501 to S503: S501: Determine the phase angle stability index based on the change range of the phase angle difference within a preset time.

[0078] Specifically, the electronic equipment calculates the maximum and minimum values ​​of the three-phase phase angle difference within a preset time period, calculates the difference between the maximum and minimum values ​​to obtain the phase angle difference change amplitude, and compares the phase angle difference change amplitude with a preset threshold. When the phase angle difference change amplitude is less than or equal to the preset threshold, the phase angle stability index is equal to the phase angle difference change amplitude divided by the preset threshold; when the phase angle difference change amplitude is greater than the preset threshold, the phase angle stability index is equal to the sum of the phase angle difference change amplitude and the portion exceeding the preset threshold divided by the preset threshold. The smaller the phase angle stability index value, the more stable the generator operation. A piecewise function is used to establish the correspondence between the phase angle difference change amplitude and the preset threshold. Within the normal operating range, linear mapping is used for normalization. Under abnormal conditions, the ratio of the excess portion is used as a penalty term, ensuring both the sensitivity of the index within the normal operating range and highlighting the early warning effect of abnormal conditions, enabling the phase angle stability index to comprehensively and accurately reflect the generator's operating status.

[0079] S502: Determine the fault sensitivity index based on the consistency between the amplitude of harmonic components and the change trend of neutral point voltage.

[0080] Specifically, the electronic equipment calculates the Pearson correlation coefficient between the three-phase harmonic component amplitude sequence and the neutral point voltage sequence, obtaining the absolute value of the correlation coefficient. The electronic equipment also calculates the direction of change of the two sequences at adjacent sampling points, and statistically analyzes the proportion of points with the same direction of change to obtain the trend overlap. Finally, the electronic equipment weights and sums the absolute value of the correlation coefficient and the trend overlap according to preset weight coefficients to obtain a fault sensitivity index, where the sum of all weight coefficients is one. A higher fault sensitivity index indicates more pronounced generator fault characteristics. This calculation method, by combining the two dimensions of correlation coefficient and trend overlap, achieves a comprehensive evaluation of the static correlation and dynamic consistency between harmonic components and the neutral point voltage. The correlation coefficient reflects the overall linear correlation of the sequences, while the trend overlap captures local variation characteristics. The weighted fusion of the two ensures the comprehensiveness of fault feature identification and provides flexibility to adapt to different operating scenarios through weight adjustment.

[0081] S503: Combine the phase angle stability index with the fault sensitivity index to obtain a comprehensive state feature vector.

[0082] Specifically, the electronic equipment normalizes the three-phase phase angle stability index to obtain steady-state operation characteristic components and normalizes the three-phase fault sensitivity index to obtain fault state characteristic components. The electronic equipment combines the steady-state operation characteristic components and the fault state characteristic components according to a preset encoding format to form a comprehensive state feature vector containing generator steady-state operation characteristics and fault characteristic information.

[0083] S103: Classify the operating conditions of the target generator to obtain multiple operating modes, and determine the corresponding health state trajectory clusters based on the comprehensive state feature vectors under each operating mode.

[0084] In this embodiment, the operating condition represents the operating characteristics of the generator under different operating conditions, and the operating condition is used to distinguish the operating modes of the generator; the health status trajectory cluster represents the trajectory set formed by the change of the comprehensive state feature vector of the generator over time under each operating mode, and the health status trajectory cluster is used to characterize the normal operating state characteristics of the generator.

[0085] Specifically, the electronic equipment classifies the operating conditions of the target generator based on active power and reactive power, dividing the generator operating conditions into heavy-load operating mode, light-load operating mode, and no-load operating mode. The electronic equipment classifies the operating modes based on the ratio of active power to rated power, and records the real-time value of the comprehensive state feature vector at a preset sampling interval for each operating mode. The feature vector sequence within the target time period constitutes the operating trajectory, and multiple operating trajectories under the same operating mode are aggregated to form a health status trajectory cluster for that operating mode, used to characterize the normal operating characteristics of the generator under that operating mode.

[0086] Based on the above embodiments, as an optional embodiment, the operating conditions of the target generator are classified to obtain multiple operating modes. Based on the comprehensive state feature vectors under each operating mode, the corresponding health state trajectory cluster is determined. This step may further include steps S601 to S603: S601: Classify the operating conditions of the target generator based on active power and reactive power to obtain multiple operating modes.

[0087] In the embodiments of this application, active power represents the effective power output by the generator to the grid, and active power is used to characterize the actual load level of the generator; reactive power represents the ineffective power output by the generator to the grid, and reactive power is used to characterize the excitation state of the generator; operating mode represents the working state of the generator under different combinations of active and reactive power.

[0088] Specifically, the electronic equipment inputs the active and reactive power of the target generator into the operating condition classification module, dividing the operating condition boundaries on a two-dimensional plane composed of active and reactive power. The electronic equipment then judges based on preset thresholds: when the actual values ​​of both active and reactive power exceed a preset high-level threshold, it is classified as a heavy-load operation mode; when both actual values ​​are below a preset low-level threshold, it is classified as an no-load operation mode; all other cases are classified as light-load operation modes. The electronic equipment maps the generator's real-time operating conditions to the corresponding operating modes.

[0089] S602: Record the continuous trajectory of the comprehensive state feature vector changing over time within each operating mode during the target time period.

[0090] In this embodiment, the continuous trajectory represents the change path of the comprehensive state feature vector on the time axis, and the continuous trajectory is used to describe the dynamic evolution process of the generator's operating state. The target time period represents the time interval for recording the generator's operating state, and the target time period is used to determine the scope of data collection for the operating data.

[0091] Specifically, the electronic device sets the target time period to the start and end of data acquisition, calculates the comprehensive state feature vector within the target time period, arranges the comprehensive state feature vector in chronological order to form a time-series data sequence, synchronously records the working status identifier of each operating mode, maps the time-series data sequence to a multi-dimensional feature space to generate a continuous trajectory, and each data point of the continuous trajectory corresponds to a comprehensive state feature vector at a certain moment.

[0092] S603: Based on the continuous trajectories of each operating mode, determine the health status trajectory clusters within each operating mode.

[0093] Specifically, the electronic device first divides the complete continuous trajectory within the target time period into trajectory segments corresponding to different operating modes based on the operating mode identifier. For each operating mode, the electronic device calculates the probability density distribution of the trajectory segment corresponding to that mode in the multidimensional feature space, counts the frequency of trajectory points appearing in each region, and constructs the probability density function of the trajectory points using the kernel density estimation method. The electronic device defines the region with a probability density exceeding a preset threshold as the boundary of the trajectory cluster under that operating mode, calculates the arithmetic mean of all trajectory points within the trajectory cluster to determine the center position of the trajectory cluster under that operating mode, and calculates the variance of the Euclidean distance from the trajectory point to the center of the trajectory cluster to obtain the dispersion of the trajectory cluster under that operating mode. The electronic device uses the boundary range, center position, and dispersion of the trajectory cluster under each operating mode as feature parameters of the corresponding health status trajectory cluster, thereby constructing the corresponding health status trajectory cluster.

[0094] For example, for a complete and continuous trajectory within a target time period, the electronic device extracts the trajectory segments belonging to the heavy-load operation mode based on the operation mode identifier. The electronic device calculates the trajectory point density of each region within this trajectory segment and uses a kernel density estimation method to obtain the probability distribution function of the trajectory points. When the probability density of a certain region exceeds a set 90% threshold, that region is defined as the boundary of the trajectory cluster for the heavy-load operation mode. The electronic device calculates the average value of all trajectory points within this boundary as the center of the trajectory cluster and calculates the variance of the Euclidean distance from all trajectory points to the center to obtain the dispersion of the trajectory cluster. Finally, the electronic device uses these calculated boundary ranges, center positions, and dispersion as feature parameters of the healthy trajectory cluster under the heavy-load operation mode, thus constituting the healthy trajectory cluster under the heavy-load operation mode.

[0095] S104: Analyze the geometric characteristics of each health state trajectory cluster, and determine the protection parameters of the relay protection device used to protect the target generator based on each geometric characteristic.

[0096] Specifically, the electronic equipment calculates the geometric center coordinates and boundary curves of each health state trajectory cluster, statistically analyzes the distribution density and variance of trajectory points within each health state trajectory cluster, maps the magnitude and distribution position of the geometric center coordinates to the starting current threshold and starting voltage threshold of the relay protection device, maps the range of the boundary curves to the protection characteristic curve range of the relay protection device, and maps the distribution density and variance to the tripping delay time of the relay protection device. The electronic equipment uses the starting current threshold, starting voltage threshold, protection characteristic curve range, and tripping delay time as protection parameters of the relay protection device.

[0097] Figure 4This diagram illustrates the construction and analysis process of a health status trajectory cluster according to an embodiment of this application. It presents the construction and analysis process of the health status trajectory cluster in this invention from an evolutionary perspective, step by step. The diagram includes three consecutive stages connected by arrows, showing the process from a single data point to the formation of an analyzable geometric model from left to right.

[0098] Figure 4 The leftmost panel displays an isolated point. This represents a single, calculated composite state feature vector at a given moment. This composite state feature vector is the basic data unit that constitutes the health state trajectory cluster.

[0099] Figure 4 The central panel displays a point cloud composed of a large number of data points. It represents the set of trajectory points formed by continuously recording and plotting the comprehensive state feature vectors at multiple moments within a target time period in the feature space. This set of points is the health state trajectory cluster defined in this application, corresponding to a specific operating mode.

[0100] Figure 4 The rightmost panel displays the results of geometric feature analysis on the formed trajectory clusters. The "+" sign at the center of the point cloud represents the calculated geometric center coordinates, i.e., the ideal healthy state; the dashed outline surrounding the point cloud represents the calculated boundary curve, i.e., the safe operating range; and isolated star-shaped points falling outside the boundary represent an identified anomaly.

[0101] In conclusion, Figure 4 This paper fully illustrates the process from obtaining a single comprehensive state feature vector, to constructing a trajectory cluster representing the healthy operating state of the generator, and then to distinguishing between normal and abnormal states by analyzing the geometric characteristics of this trajectory cluster (such as geometric center coordinates and boundary curves). The results of this analysis provide direct quantitative basis for determining various protection parameters of the relay protection device (such as starting threshold and protection characteristic curve range) in subsequent steps.

[0102] Based on the above embodiments, as an optional embodiment, the geometric characteristics of each health state trajectory cluster are analyzed, and the protection parameters of the relay protection device for protecting the target generator are determined based on each geometric characteristic. This step may further include steps S701 to S706: S701: Calculate the geometric center coordinates and boundary curves of each health state trajectory cluster.

[0103] Specifically, the electronic device performs an arithmetic mean operation on the coordinates of all trajectory points within the health status trajectory cluster, calculates the mean of the trajectory points in each feature dimension, uses the vector composed of the means of each feature dimension as the geometric center coordinates of the health status trajectory cluster, calculates the distance from all trajectory points in the trajectory cluster to the geometric center, statistically analyzes the spatial distribution density of the trajectory points, uses density clustering to identify the distribution boundary of the trajectory points, and fits the distribution boundary into a continuous and smooth boundary curve, which encloses the outer contour of the health status trajectory cluster.

[0104] S702: Statistically analyze the distribution density and variance of trajectory points within each health status trajectory cluster.

[0105] Specifically, the electronic device divides the feature space into multiple grid cells, calculates the number of trajectory points in each grid cell and divides it by the volume of the grid cell to obtain the local distribution density, calculates the average local distribution density of all grid cells to obtain the overall distribution density of the healthy trajectory cluster, calculates the Euclidean distance from all trajectory points in the trajectory cluster to the geometric center, and divides the sum of the squares of the Euclidean distances by the total number of trajectory points to obtain the variance of the healthy trajectory cluster. The distribution density and variance together describe the spatial distribution characteristics of the trajectory cluster.

[0106] S703: Determine the starting current threshold and starting voltage threshold of the relay protection device used to protect the target generator based on the magnitude and distribution of the geometric center coordinates.

[0107] In this embodiment, the starting current threshold represents the current reference value for determining when the relay protection device starts operating, used to detect whether the generator current exceeds the normal operating range; the starting voltage threshold represents the voltage reference value for determining when the relay protection device starts operating, used to detect whether the generator voltage exceeds the normal operating range. The geometric center coordinates represent the center point of the health state trajectory cluster in the feature space, used to locate the core area of ​​the generator's normal operating state.

[0108] Specifically, the electronic device reads the geometric center coordinates of each health status trajectory cluster and projects these coordinates from the multi-dimensional feature space onto the current and voltage characteristic axes using a coordinate projection method. The projected values ​​are then used as the corresponding starting current and starting voltage reference values. The electronic device calculates the standard deviation from the geometric center coordinates to each characteristic axis, using this standard deviation as an index of the distribution's dispersion. Based on the magnitude of the dispersion, a corresponding mapping coefficient is determined. In one specific embodiment, this mapping coefficient is proportional to the dispersion (e.g., the standard deviation σ). For example, the mapping coefficient k = α * σ, where α is a preset empirical constant used to adjust the sensitivity of the threshold. Finally, the mapping coefficient is multiplied by the corresponding reference values ​​to obtain the final starting current and starting voltage thresholds, which serve as the protection parameters of the relay protection device.

[0109] S704: Determine the range of the protection characteristic curve of the relay protection device based on the range of the boundary curve. The range of the protection characteristic curve characterizes the safe operating range of the target generator's operating parameters.

[0110] In this embodiment of the application, the protection characteristic curve range represents the boundary of the area by which the relay protection device judges whether the generator's operating state is normal. The protection characteristic curve range is used to divide the generator's safe operating area and fault area.

[0111] Specifically, the electronic equipment reads the boundary curve equations of each health status trajectory cluster, maps the projection range of the boundary curves in the current characteristic dimension and voltage characteristic dimension to the protection characteristic curve range of the relay protection device, calculates the values ​​of the boundary curves under different combinations of operating parameters and maps them to the shape of the protection characteristic curve, sets the area after the boundary curve is mapped to the protection characteristic curve, and the protection characteristic curve range limits the safe operating area of ​​the generator operating parameters. When the operating parameters exceed the protection characteristic curve range, the relay protection device is triggered to operate.

[0112] For example, the boundary curve equation of the health status trajectory cluster acquired by an electronic device under heavy-load operation mode can be expressed as I 2 / In 2 +U 2 / Un 2 = 1, where I represents the operating current, U represents the operating voltage, and In and Un represent the rated current and rated voltage, respectively. Based on this boundary curve equation, the electronic equipment extracts the operating current range [0.8In, 1.2In] and the operating voltage range [0.9Un, 1.1Un] as the reference region. The electronic equipment analyzes the elliptical characteristics of the boundary curve and constructs the corresponding protection characteristic curve shape so that the protection characteristic curve can completely enclose the reference region, thereby accurately defining the safe operating range of the generator under heavy load operation mode.

[0113] S705: Determine the tripping delay time of the relay protection device based on the distribution density and variance.

[0114] In this embodiment, the trip delay time represents the time interval between the delay in action of the relay protection device after detecting a fault, which is used to avoid malfunctions caused by fluctuations in the normal operation of the generator.

[0115] Specifically, the electronic device reads the distribution density and variance values ​​of each health status trajectory cluster, calculates the ratio of the distribution density to the preset baseline density to obtain the density ratio coefficient, calculates the ratio of the variance to the preset baseline variance to obtain the variance ratio coefficient, multiplies the density ratio coefficient and the variance ratio coefficient to obtain the time coefficient, and multiplies the time coefficient to the preset baseline delay to obtain the tripping delay time of the relay protection device. The higher the distribution density, the longer the tripping delay time; the higher the variance, the longer the tripping delay time. The tripping delay time is used to filter fluctuations in the normal operation of the generator. The step of multiplying the density ratio coefficient and the variance ratio coefficient to obtain the time coefficient uses multiplication instead of addition, which amplifies the dual stabilization effect: when the trajectory cluster is both dense (high density) and stable (low variance), the time coefficient will decrease significantly, shortening the delay; conversely, when the trajectory cluster is both sparse and diffuse, the time coefficient will increase significantly, lengthening the delay, thus achieving a stronger filtering effect.

[0116] S706: The starting current threshold, starting voltage threshold, protection characteristic curve range, and trip delay time of the relay protection device shall be used as the protection parameters of the relay protection device used to protect the target generator.

[0117] Specifically, the electronic equipment uses the calculated starting current threshold and starting voltage threshold as the starting criteria for the relay protection device, the range of the protection characteristic curve as the operating area criterion for the relay protection device, and the tripping delay time as the operating time criterion for the relay protection device. The electronic equipment combines the starting criteria, operating area criterion, and operating time criterion into a complete protection parameter configuration scheme. The relay protection device uses the protection parameter configuration scheme to monitor the generator operating status and performs protection actions to isolate the fault when an abnormal status is detected.

[0118] The following describes an exemplary electronic device for self-matching relay protection parameters provided in an embodiment of this application. Figure 5 This is an exemplary hardware structure diagram of an electronic device for self-matching relay protection parameters provided in an embodiment of this application.

[0119] In some embodiments, the relay protection parameter self-configuration electronic device is a computer device, or the user behavior-based relay protection parameter self-configuration electronic device includes a computer device. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device stores data. The network interface of the computer device is used to communicate with other external terminals or servers via a network connection. In some embodiments, the network interface can be a wired network interface; in some embodiments, the network interface can also be a wireless network interface. When the computer program is executed by the processor, it implements the methods in the embodiments of this application.

[0120] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0121] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0122] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0123] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0124] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A method for self-matching relay protection parameters, characterized in that, The method includes: Obtain the three-phase voltage, three-phase current, and neutral point voltage of the target generator within a preset time period; The phase angle difference is determined based on the three-phase voltage and three-phase current, and a comprehensive state feature vector is constructed based on the phase angle difference and the neutral point voltage. The operating conditions of the target generator are classified to obtain multiple operating modes. Based on the comprehensive state feature vector of each operating mode, the corresponding health state trajectory cluster is determined. Analyze the geometric features of each health state trajectory cluster, and determine the protection parameters of the relay protection device for protecting the target generator based on each geometric feature; The operating conditions of the target generator are classified into multiple operating modes. Based on the comprehensive state feature vector under each operating mode, a corresponding health state trajectory cluster is determined, specifically including: The operating conditions of the target generator are classified based on active power and reactive power to obtain multiple operating modes; Within the target time period, record the continuous trajectory of the comprehensive state feature vector changing over time in each of the aforementioned operating modes; Based on the continuous trajectories of each of the aforementioned operating modes, a cluster of health status trajectories within each operating mode is determined; Specifically, the process includes analyzing the geometric characteristics of each health state trajectory cluster and determining the corresponding protection parameters of the relay protection device for protecting the target generator based on these geometric characteristics. Calculate the geometric center coordinates and boundary curves of each health state trajectory cluster; Statistically analyze the distribution density and variance of trajectory points within each health state trajectory cluster; The starting current threshold and starting voltage threshold of the relay protection device used to protect the target generator are determined based on the magnitude and distribution of the geometric center coordinates. The range of the protection characteristic curve of the relay protection device is determined based on the range of the boundary curve, and the range of the protection characteristic curve characterizes the safe operating range of the target generator's operating parameters; The tripping delay time of the relay protection device is determined based on the distribution density and variance. The starting current threshold, starting voltage threshold, protection characteristic curve range, and trip delay time of the relay protection device are used as protection parameters for the relay protection device used to protect the target generator.

2. The relay protection parameter self-matching method according to claim 1, characterized in that, The process of determining the phase angle difference based on the three-phase voltage and three-phase current, and constructing a comprehensive state feature vector based on the phase angle difference and the neutral point voltage, specifically includes: Dynamic window sliding sampling is performed on the three-phase voltage and the three-phase current to obtain the voltage and current sampling sequence within a preset sampling period; The voltage and current sampling sequence is subjected to wavelet transform to obtain a time-frequency distribution feature map, and the fundamental frequency component and harmonic component of the three-phase voltage and three-phase current are extracted from the time-frequency distribution feature map respectively. Based on each of the fundamental frequency components, the phase angle difference between the three-phase voltage and the three-phase current is determined; A comprehensive state feature vector is constructed based on the phase angle difference, the amplitude of the harmonic components, and the neutral point voltage.

3. The relay protection parameter self-matching method according to claim 2, characterized in that, The determination of the phase angle difference between the three-phase voltage and the three-phase current based on each of the fundamental frequency components specifically includes: Perform a Hilbert transform on each of the fundamental frequency components to obtain the corresponding orthogonal components; Based on each fundamental frequency component and its corresponding quadrature component, the phase angles of the three-phase voltage and three-phase current are determined respectively. Based on each of the aforementioned phase angles, the initial phase difference between the three-phase voltage and the three-phase current is determined; The initial phase difference is subjected to a moving average filter to obtain the phase angle difference between the three-phase voltage and the corresponding three-phase current.

4. The relay protection parameter self-matching method according to claim 2, characterized in that, The construction of a comprehensive state feature vector based on the phase angle difference, the amplitude of the harmonic components, and the neutral point voltage specifically includes: Based on the variation range of the phase angle difference within a preset time period, a phase angle stability index is determined; Based on the consistency between the amplitude of the harmonic components and the changing trend of the neutral point voltage, the fault sensitivity index is determined. The phase angle stability index and the fault sensitivity index are combined to obtain a comprehensive state feature vector.

5. The relay protection parameter self-matching method according to claim 2, characterized in that, The step of performing wavelet transform on the voltage and current sampling sequence to obtain a time-frequency distribution feature map, and extracting the fundamental frequency component and harmonic component of the three-phase voltage and three-phase current from the time-frequency distribution feature map, specifically includes: The voltage and current sampling sequences of the three-phase voltage and three-phase current are subjected to wavelet transform hierarchical decomposition to obtain wavelet coefficients at different scales; Based on the wavelet coefficients, a time-frequency distribution feature map reflecting the change of the frequency components of the three-phase voltage and three-phase current over time is obtained; In the time-frequency distribution feature map, the energy distribution ratio of each frequency band is calculated, and the frequency band with the largest energy ratio is determined as the fundamental frequency component. In the time-frequency distribution feature map, other frequency bands outside the frequency band where the fundamental frequency component is located are identified as harmonic components.

6. An electronic device with self-matching relay protection parameters, characterized in that, The electronic device includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the electronic device to perform the method as described in any one of claims 1-5.

7. A computer program product containing instructions, characterized in that, When the computer program product is run on an electronic device with self-matched relay protection parameters, the electronic device performs the method as described in any one of claims 1-5.

8. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on an electronic device with self-matched relay protection parameters, the electronic device performs the method as described in any one of claims 1-5.

Images