Intelligent nuclear phase method and system for transformer substation

Abstract

The invention belongs to the technical field of intelligent substations, and particularly relates to an intelligent nuclear phase method and system for a substation. According to the method, a dynamic priority evaluation model is constructed, multi-source data of four kinds of basic factors including voltage phase difference, environment temperature and humidity, equipment vibration and historical errors are synthesized, and a fuzzy comprehensive evaluation method is utilized to calculate weights and dynamically sort the weights; a multi-port phase scattering model is constructed based on transformer substation topology, impedance mismatch points are analyzed by means of a Smith chart, and dual-stage optimization is implemented; constructing a three-layer decision input set according to the optimization model, designing a fuzzy logic driven adaptive rule engine, dynamically outputting a working condition level and matching a nuclear phase strategy; multi-dimensional quality evaluation, dynamic updating of factor weights and optimization of fuzzy model parameters are carried out by monitoring indexes such as reflection coefficients in real time, abnormal mode recognition and compensation strategy solidification are realized by fusing LSTM and a graph neural network, a self-evolution closed-loop system is formed, and finally, crossing of a nuclear phase process from accurate matching to autonomous evolution is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent phase verification in substations, and particularly relates to an intelligent phase verification method and system for substations. Background Art

[0002] In the safe and stable operation of the power system, the phase verification operation of intelligent substations is a crucial task. The accuracy of phase verification is directly related to the grid connection of the power system, the commissioning of equipment, and the reliable operation of the system. With the rapid development of the smart grid, the degree of intelligence of substations has been continuously improved, and their operating environments and equipment characteristics have become increasingly complex, which poses higher requirements for phase verification technology.

[0003] Traditional intelligent phase verification methods for substations have many deficiencies when facing modern complex substation environments. On the one hand, traditional methods often fail to fully consider the interference of various influencing factors on the phase verification results. For example, as a core parameter in phase verification, the accuracy of the voltage phase difference factor is affected by various factors, but traditional technologies may not conduct a comprehensive dynamic evaluation and compensation for it.

[0004] In terms of the construction of the phase verification model, there are obvious defects in the existing technologies. Most of the existing phase verifications are based on simple electrical connection relationships, and fail to fully consider the complexity of the substation topology structure and the phase response characteristics of each electrical connection port. At the same time, the existing models also have deficiencies in dealing with impedance matching problems. Since there is no effective method to identify impedance mismatch points, the phase verification model cannot be optimized specifically, resulting in problems such as phase reflection and offset during the actual phase verification process, seriously affecting the accuracy of phase verification. Moreover, in terms of the execution of phase verification strategies and resource scheduling, the existing technologies lack flexibility and adaptability. The existing phase verification strategies often adopt fixed execution modes and cannot dynamically adjust the phase verification methods according to the actual working conditions. Finally, for the errors generated during the phase verification process, the compensation mechanism of traditional technologies is not perfect. Most of the existing error compensation methods adopt fixed compensation coefficients and cannot be dynamically adjusted according to the priorities and change characteristics of the basic factors. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent phase verification method and system for substations. This method integrates dynamic priority evaluation, multi-port phase scattering model, and closed-loop feedback to achieve adaptive optimization of phase verification strategies and self-evolution of the system, and improve the phase matching accuracy and stability of intelligent substations.

[0006] The technical solution of the present invention is as follows:

[0007] One of the technical solutions of the present invention is to provide an intelligent phase verification method for substations, including:

[0008] Taking the voltage phase difference factor, ambient temperature and humidity factor, equipment vibration factor, and historical error factor as basic factors, judge the dynamic priorities of the basic factors by constructing a dynamic priority evaluation model;

[0009] Based on the substation topology, abstract the signal transmission and reflection characteristics of each electrical connection port into a multi-port phase scattering model; use the Smith chart to analyze the port phase scattering characteristics, identify the impedance mismatch points, and implement a two-stage optimization for the multi-port phase scattering model;

[0010] Execute the phase comparison strategy, and construct a decision input set based on the optimized phase scattering model; based on the fuzzy logic and interface selection idea, design an adaptive rule engine, dynamically output the phase comparison working condition level, and automatically select the phase comparison strategy accordingly;

[0011] Conduct multi-dimensional quality evaluation by real-time monitoring of the port phase matching degree and reflection coefficient, and dynamically update the factor weights and optimize the fuzzy model parameters through the evaluation results to form a self-evolving system.

[0012] As a further option of this method, the dynamic priority evaluation model uses the fuzzy comprehensive evaluation method to calculate the weights of each basic factor, including:

[0013] Establish a triangular membership function, Gaussian membership function or S-shaped membership function, and map the characteristic values of each basic factor to the interval [0,1];

[0014] Define the fuzzy weight vector, and calculate the priority index through the formula where is the weight of the th factor, is the membership degree of the th factor, is the membership degree of the th factor;

[0015] Based on the priority index, dynamically sort the voltage phase difference factor, ambient temperature and humidity factor, equipment vibration factor and historical error factor, and drive the adjustment of the phase comparison strategy.

[0016] As a further option of this method, abstract the substation electrical equipment into a multi-port network, and define the scattering matrix as a × complex matrix, where the element represents the scattering parameter of port to port ;

[0017] Calculate the scattering parameter through the formula where, is the scattering parameter, For port The reflected wave, For port The reflected wave of the incident wave;

[0018] The multi-port phase scattering model is weightedly optimized by combining the factor priorities output by the dynamic priority evaluation model.

[0019] As a further option of the method, the two-stage optimization includes:

[0020] First-stage optimization: For high-priority ports, impedance parameters are dynamically adjusted using parallel or series reactance elements. For low-priority ports, a fixed compensation strategy is used, with compensation values calculated based on historical error factor feedback.

[0021] Second stage optimization: Optimize the objective function through full-site collaboration Achieve global optimal matching, where is the impedance vector of all ports, Frequency The phase shift, is the reflection coefficient weight coefficient, is the operating frequency band of the substation, For the Port frequency The modulus of the reflection coefficient.

[0022] As a further option of this method, the decision input set includes three layers of data:

[0023] Basic layer: including phase offset, reflection coefficient modulus, and impedance matching status;

[0024] State layer: includes dynamic priority indicators of basic factors output by the dynamic priority evaluation model;

[0025] Strategy layer: including compensation coefficients and priority weights driven by historical errors;

[0026] The data are normalized and mapped to a uniform dimension interval.

[0027] As a further option of this method, the adaptive rule engine includes:

[0028] Construct a fuzzy logic rule base, define the input variables as phase offset, reflection coefficient, and basic factor dynamic priority index, and divide the fuzzy intervals into "high / medium / low";

[0029] Dynamically match the core phase interface type based on fuzzy reasoning results. High-speed synchronous interface is enabled for optimal working conditions, low-bandwidth interface is switched to for intermediate working conditions, and redundant interface is activated for poor working conditions.

[0030] Output core phase execution strategy, including full-node parallel calibration of superior strategy, historical error sorting polling of intermediate strategy, and dual-channel redundant verification of poor strategy.

[0031] As a further option of this method, the multi-dimensional quality assessment includes:

[0032] A three-level quality indicator system is defined: base-layer indicators include port reflection coefficient modulus, phase offset, and impedance normalization deviation; status-layer indicators include dynamic priority weight volatility, historical error recurrence probability, and equipment vibration anomaly; and performance-layer indicators include phase strategy resource utilization, interface switching response delay, and site-wide collaborative optimization convergence speed.

[0033] Perform weighted analysis on real-time data through a sliding window algorithm to generate a multi-dimensional quality assessment report.

[0034] As a further option of this method, the factor weight dynamic update mechanism includes:

[0035] Normalize the voltage phase difference, temperature and humidity, vibration amplitude and historical error data collected in real time;

[0036] Adjust the fuzzy interval parameters of the triangle membership function 、 and ,in 、 and are fuzzy interval parameters, representing the lower limit, optimal value and upper limit of the factor eigenvalue respectively;

[0037] Introducing a time decay factor , where 0< <1, by formula Update the weights, where For the Moment The fuzzy weight vector of the basic factors, is a quality assessment indicator, For the The eigenvalues of the basic factors.

[0038] As a further option of this method, the fuzzy model parameter optimization includes:

[0039] The kernel density estimation method is used to dynamically divide the fuzzy interval;

[0040] Assign dynamic confidence coefficients to fuzzy rules, which are calculated jointly by the historical rule triggering accuracy and the current quality assessment deviation rate;

[0041] Genetic algorithm is introduced to optimize the shape parameters of fuzzy membership function;

[0042] When multiple fuzzy rules conflict, the DS evidence theory is used to fuse multi-source evidence and output the working condition level judgment with the maximum comprehensive confidence.

[0043] As a further option of this method, the abnormal pattern recognition and compensation strategy generation steps in the self-evolution system include:

[0044] Long short-term memory networks are used to capture the long-term dependencies of voltage phase difference and equipment vibration;

[0045] Modeling electrical coupling relationships in substation topology using graph neural networks;

[0046] The identified abnormal patterns are labeled and stored, and the compensation solutions that recur frequently are solidified into new fuzzy rules.

[0047] As a further option of this method, the voltage phase difference factor is obtained by Calculate; where, is the voltage phase difference factor, is the actual measured voltage phase difference, is the maximum voltage phase difference allowed by the system;

[0048] The environmental temperature and humidity factors are Calculate; where, is the environmental temperature and humidity factor, and are the weights of temperature and humidity, and are functions of the effects of temperature and humidity on device performance;

[0049] The equipment vibration factor is Calculate; where, is the equipment vibration factor, and are the weights of vibration frequency and amplitude respectively, and are functions of the effects of vibration frequency and amplitude on equipment performance;

[0050] The historical error factor is based on Calculate; where, is the historical error factor, For the The weight of the historical error, For the The severity of the historical error.

[0051] As a further option of this method, the core phase execution strategy includes:

[0052] Superior strategy: Rely on high-precision timestamps to achieve full-port parallel calibration, and prioritize ensuring the matching speed of the main transformer and bus through a load balancing algorithm;

[0053] Intermediate strategy: Allocate limited bandwidth resources based on the time slice rotation strategy, monitor the node progress in real time, and dynamically release computing resources;

[0054] Inferior strategy: Independently phase-check the main and backup channels, trigger a secondary optimization process when the comparison result deviation occurs, automatically switch to the backup channel in case of anomalies, and notify the operation and maintenance personnel.

[0055] The second technical solution of the present invention is to provide a substation intelligent phase-checking method system, including:

[0056] A dynamic priority evaluation module, which uses the voltage phase difference factor, ambient temperature and humidity factor, equipment vibration factor, and historical error factor as basic factors, and judges the dynamic priority of the basic factors by constructing a dynamic priority evaluation model;

[0057] A multi-port modeling and optimization module, which abstracts the signal transmission and reflection characteristics of each electrical connection port into a multi-port phase scattering model based on the substation topology; analyzes the port phase scattering characteristics using a Smith chart, identifies impedance mismatch points, and implements two-stage optimization on the multi-port phase scattering model;

[0058] A phase-checking strategy execution module, which is used to execute the phase-checking strategy, construct a decision input set based on the optimized phase scattering model; design an adaptive rule engine based on fuzzy logic and interface selection ideas, dynamically output the phase-checking working condition level, and automatically select the phase-checking strategy accordingly;

[0059] A quality evaluation and optimization module, which conducts multi-dimensional quality evaluation by real-time monitoring of the port phase matching degree and reflection coefficient, dynamically updates the factor weights and optimizes the fuzzy model parameters through the evaluation results, and forms a self-evolving system.

[0060] The beneficial effects brought by the technical solution provided by the embodiments of the present application at least include the following beneficial effects:

[0061] This technical solution realizes a breakthrough innovation in the phase verification technology of intelligent substations by constructing a dynamic priority evaluation model, a multi-port phase scattering model, and a closed-loop feedback self-evolution system. Its core advantages are as follows: The multi-factor dynamic priority evaluation model deeply integrates multi-dimensional data such as voltage phase difference, ambient temperature and humidity, equipment vibration, and historical errors. Combining the fuzzy comprehensive evaluation method with the triangular membership function, it dynamically adjusts the weights of each factor, breaks through the limitations of traditional static models, and significantly improves the adaptability under complex working conditions; The multi-port phase scattering model and two-stage optimization accurately identify impedance mismatch points through the Smith chart. Combining single-port basic matching with full-station collaborative optimization, it realizes the minimization of phase shift within a wide frequency band and the global optimum of reflection loss; The adaptive phase verification strategy and the fuzzy logic rule engine dynamically output excellent / medium / poor working condition levels based on a three-layer decision input set, and differentially match high-speed synchronous interfaces, low-bandwidth time-sharing scheduling, or redundant channel verification strategies, taking into account both efficiency and fault tolerance; The closed-loop feedback system drives the update of factor weights and the optimization of fuzzy model parameters through multi-dimensional quality evaluation, and introduces deep learning technology to solidify high-frequency compensation schemes as new rules, realizing system self-learning and fault prediction.

[0062] This technical solution has been piloted in multiple intelligent substations, verifying its significant application effect and practical value. In terms of improving phase verification accuracy and efficiency, through multi-port collaborative optimization, the modulus of the port reflection coefficient is reduced below the threshold, the maximum phase shift of the whole station is controlled within the allowable range, the parallel calibration speed of all nodes under excellent working conditions is increased by more than 30%, the utilization rate of limited bandwidth under medium working conditions is increased by 25%, and the fault switching response time under poor working conditions is shortened to the millisecond level. In terms of enhancing environmental adaptability and fault tolerance, the dynamic priority evaluation model can respond in real time to environmental coupling effects such as temperature and humidity fluctuations and vibration anomalies, and still maintain phase verification stability under extreme weather. The dual-channel redundancy verification mechanism of the poor-level strategy effectively isolates hardware faults and ensures the continuous operation of the power grid. In terms of intelligent and standardized integration, the output of MMS service messages compliant with the IEC61850 standard is seamlessly connected to the SCADA system, and the full-life cycle health management module generates digital portraits and efficiency heat maps, providing data-driven decision-making support for digital operation and maintenance.

[0063] This technical solution constructs a complete technical system for the phase verification technology of intelligent substations through the four-dimensional coordination of "multi-factor dynamic priority modeling - multi-port impedance optimization - adaptive strategy execution - closed-loop self-evolution", and solves the bottleneck problems of traditional methods in terms of environmental adaptability, resource allocation efficiency, and long-term stability. Its innovation is reflected in:

[0064] The dynamic priority model breaks through the limitations of static weights and realizes the non-linear coupling analysis of multiple factors;

[0065] The two-stage optimization algorithm takes into account both single-port matching and full-station coordination, improving the wide-frequency band performance;

[0066] The hierarchical strategy of the fuzzy logic rule engine matches resource constraints to ensure the priority of key nodes;

[0067] The self-evolving system realizes equipment aging compensation and fault mode self-learning through a data-driven closed loop.

[0068] This technology has been piloted in multiple intelligent substations, verifying its significant advantages in improving the accuracy of phase comparison, reducing operation and maintenance costs, and enhancing system robustness, providing a replicable technical path for the intelligent upgrade of the new power system. Description of the Drawings

[0069] Figure 1 It is a schematic diagram of the overall process of the intelligent phase comparison method for substations;

[0070] Figure 2 It is a detailed flowchart of step S100 of the intelligent phase comparison method for substations;

[0071] Figure 3 It is a detailed flowchart of step S200 of the intelligent phase comparison method for substations;

[0072] Figure 4 It is a detailed flowchart of step S300 of the intelligent phase comparison method for substations;

[0073] Figure 5 It is a detailed flowchart of step S400 of the intelligent phase comparison method for substations. Detailed Implementation Modes

[0074] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0075] Traditional intelligent phase comparison technologies for substations have many technical bottlenecks when facing the complex operating environment and high-precision phase comparison requirements of modern intelligent substations. In order to improve the accuracy, reliability, and adaptability of phase comparison, there is an urgent need for a new phase comparison method that can comprehensively consider multi-source data, dynamically evaluate the priority of basic factors, optimize the phase comparison model, intelligently schedule resources, and achieve precise error compensation. Please refer to Figure 1 , which shows an intelligent phase comparison method for substations provided by an embodiment of the present invention. The method includes:

[0076] S100: Taking the voltage phase difference factor, ambient temperature and humidity factor, equipment vibration factor, and historical error factor as basic factors, and judging the dynamic priority of the basic factors by constructing a dynamic priority evaluation model;

[0077] S200: Based on the substation topology, abstract the signal transmission and reflection characteristics of each electrical connection port into a multi-port phase scattering model; use the Smith chart to analyze the port phase scattering characteristics, identify impedance mismatch points, and implement a two-stage optimization for the multi-port phase scattering model;

[0078] S300: Execute the phase verification strategy, construct a decision input set based on the optimized phase scattering model; design an adaptive rule engine based on fuzzy logic and interface selection ideas, dynamically output the phase verification working condition level, and automatically select the phase verification strategy accordingly;

[0079] S400: Conduct multi-dimensional quality assessment by real-time monitoring of the port phase matching degree and reflection coefficient, dynamically update the factor weights and optimize the fuzzy model parameters through the evaluation results to form a self-evolving system;

[0080] The specific solution is as follows:

[0081] In the substation intelligent phase verification method, S100 involves taking multiple influencing factors as basic factors, and judging the dynamic priorities of these basic factors through a dynamic evaluation model of multi-source data fusion and environmental coupling.

[0082] Please refer to Figure 2 , which shows a flowchart of an exemplary substation intelligent phase verification method S100 of the present application, and its content includes:

[0083] S110: Obtain multi-source heterogeneous data and calculate multi-dimensional basic factor data;

[0084] The multi-dimensional basic factors include: voltage phase difference factor, environmental temperature and humidity factor, equipment vibration factor, and historical error factor. The data corresponding to the multi-dimensional basic factors include: voltage phase difference data, environmental temperature and humidity data, equipment vibration data, and historical error data.

[0085] In a possible implementation manner, the voltage phase difference data is obtained through a voltage transformer and a current transformer. The environmental temperature and humidity data is obtained through a temperature and humidity sensor. The equipment vibration data is obtained through an acceleration sensor or a vibration sensor. The historical error data is extracted from the historical record database;

[0086] Calculate the voltage phase difference factor, environmental temperature and humidity factor, equipment vibration factor, and historical error factor based on the voltage phase difference data, environmental temperature and humidity data, equipment vibration data, and historical error data.

[0087] In a possible implementation manner, the voltage phase difference factor is calculated by the following formula:

[0088] ;

[0089] Among them, is the voltage phase difference factor, is the actually measured voltage phase difference, is the maximum allowable voltage phase difference of the system.

[0090] In a possible implementation, the environmental temperature and humidity factor is calculated by comprehensively considering the influences of temperature and humidity:

[0091] ;

[0092] Among them, is the environmental temperature and humidity factor, and are the weights of temperature and humidity respectively, and are the functions of the influences of temperature and humidity on the device performance respectively.

[0093] In a possible implementation, the device vibration factor is calculated by the following formula:

[0094] ;

[0095] Among them, is the device vibration factor, and are the weights of vibration frequency and amplitude respectively, and are the functions of the influences of vibration frequency and amplitude on the device performance respectively.

[0096] In a possible implementation, the historical error factor is calculated according to the error frequency and severity in historical data:

[0097] ;

[0098] Among them, is the historical error factor, is the weight of the th historical error, is the severity of the th historical error.

[0099] S120: Construct a dynamic priority evaluation model.

[0100] By constructing a dynamic priority evaluation model, to determine the priorities of factors such as voltage phase difference, environmental temperature and humidity, device vibration, and historical error in the phase comparison process. The core of the dynamic priority evaluation model lies in integrating multi-source data and dynamically adjusting the weights of each factor based on the environmental coupling effect, so as to improve the adaptability and accuracy of the phase comparison strategy;

[0101] The dynamic priority evaluation model adopts a multi-factor weighted comprehensive evaluation method, where the weights of each factor are calculated by the fuzzy comprehensive evaluation method. The basic idea of this method is to use the theory of fuzzy mathematics to combine qualitative analysis with quantitative calculation to more reasonably reflect the influence of each factor under different environmental conditions. The input of the model is the characteristic indexes of each factor, and the output is the priority weight of each factor;

[0102] Specifically, let represent four basic factors, namely the voltage phase difference factor, the environmental temperature and humidity factor, the equipment vibration factor, and the historical error factor. The corresponding characteristic vectors are , where represents the th quantization characteristic value of the basic factor. In order to calculate the weight of each factor, it is first necessary to establish a fuzzy membership function to measure the influence degree of the characteristic value on the phase comparison process;

[0103] The form of the fuzzy membership function is selected according to actual needs, including triangular membership function, Gaussian membership function or S-shaped membership function.

[0104] Exemplarily, using the triangular membership function, the membership degree of the th basic factor is expressed as:

[0105] ;

[0106] where , and are fuzzy interval parameters, representing the lower limit, optimal value and upper limit of the factor characteristic value respectively, represents the th quantization characteristic value of the basic factor. Through the membership function, the characteristic values of each factor can be mapped to the interval [0,1] to measure its influence degree on the phase comparison process.

[0107] After determining the membership degree of each factor, the fuzzy comprehensive evaluation method is used to calculate the weight of each basic factor. Let the fuzzy weight vector be , and the calculation formula of the fuzzy comprehensive evaluation is:

[0108] ;

[0109] where is the weight of the th factor, satisfying ; is the membership degree of the th factor, is the th characteristic value of the factor;

[0110] The priority index reflects the overall impact of each factor on the phase verification process in the current environment and can be used to dynamically adjust the phase verification strategy. The voltage phase difference factor, ambient temperature and humidity factor, equipment vibration factor, and historical error factor are dynamically sorted through the priority index;

[0111] In the intelligent substation phase verification method, based on the substation topology, S200 abstracts the signal transmission and reflection characteristics of each electrical connection port into a multi-port phase scattering model; uses the Smith chart to analyze the port phase scattering characteristics, identifies the impedance mismatch points, and implements a two-stage optimization of the multi-port phase scattering model.

[0112] Please refer to Figure 3 , which shows the flowchart of an exemplary intelligent substation phase verification method S200 of the present application, and its content includes:

[0113] S210: Construction of the multi-port phase scattering model.

[0114] In the process of intelligent substation phase verification, the construction of the multi-port phase scattering model is the basis for achieving high-precision phase matching. Based on the basic factors in the priority order and the substation topology, the phase response characteristics of each electrical connection port are abstracted into a multi-port phase scattering model, and the phase response characteristics of each port are described based on the scattering parameters (S-parameters).

[0115] In a possible implementation, the electrical equipment in the substation is regarded as a multi-port network, and each port corresponds to an electrical connection point. Assuming the substation contains ports, then its scattering matrix is a × complex matrix, where the element represents the scattering parameter of port to port , describes the transmission and reflection characteristics of the signal between the ports, and is the key to analyzing the phase mismatch.

[0116] The definition of the multi-port phase scattering model is as follows:

[0117] ;

[0118] where is the scattering parameter, is the reflected wave of port , is the incident wave of the reflected wave of port .

[0119] The construction of the multi-port phase scattering model provides a mathematical basis for subsequent Smith chart analysis and optimization, enabling the phase verification process to accurately identify impedance mismatch points and implement dynamic adjustment.

[0120] S220: Analyze the port phase scattering characteristics using the Smith chart to identify impedance mismatch points.

[0121] Based on the multi-port phase scattering model, Smith chart analysis is used to identify impedance mismatch points for each port. The Smith chart is a polar coordinate plot for visualizing complex impedance, which can intuitively display the normalized impedance characteristics of the port and its deviation from the ideal matching state.

[0122] In one possible implementation, the operating steps for analyzing the port phase scattering characteristics using the Smith chart to identify impedance mismatch points include:

[0123] S221: Calculate the normalized impedance and reflection coefficient.

[0124] For any port , its normalized impedance is defined as: ; where is the actual impedance of the port, is the system characteristic impedance, usually 50Ω or 75Ω. After normalization, the impedance value is mapped inside the unit circle of the Smith chart for intuitive analysis.

[0125] The reflection coefficient is a key parameter for measuring the impedance matching degree of the port, and its calculation formula is: . The modulus value of the reflection coefficient represents the proportion of reflected energy at the port. The closer its value is to 0, the better the matching. On the Smith chart, the point corresponding to the reflection coefficient is closer to the center of the circle, indicating a higher impedance matching degree.

[0126] S222: Analyze the reflection coefficients of all ports.

[0127] Assume the substation contains ports, then the reflection coefficient vector can be used to quantify the matching state of each port. The determination criterion for mismatch points is usually based on the threshold of the modulus value of the reflection coefficient:

[0128] ;

[0129] where is the preset reflection coefficient threshold. If the reflection coefficient of a certain port exceeds this threshold, it is considered that there is a significant impedance mismatch at this port, and this point is marked as a mismatch point.

[0130] Through Smith chart analysis, impedance mismatch points for each port can be efficiently identified, providing an accurate adjustment direction for subsequent two-stage optimization.

[0131] S230: Based on the identified mismatch points, perform a two-stage optimization on the multi-port phase scattering model.

[0132] After identifying the impedance mismatch points, the goal of the first-stage optimization is to minimize the phase reflection coefficient of a single port by adjusting the impedance parameters of each port, thereby achieving basic impedance matching.

[0133] In a possible implementation, for high-priority ports, dynamic adjustment is usually achieved by shunting or series-connected reactance elements to reduce the reflection coefficient. Exemplarily, by adding adjustable reactance elements to change the reactance part of the port, gradually approaching the optimal matching state.

[0134] In a possible implementation, for low-priority ports, due to resource limitations, a fixed compensation strategy is usually adopted. Exemplarily, a fixed capacitor or inductor is shunted at the port. At this time, the calculation of the compensation value needs to be based on the feedback of the historical error factor.

[0135] Finally, after the first-stage optimization is completed, the modulus of the reflection coefficient of each port should be lower than the preset threshold to ensure the realization of basic impedance matching.

[0136] After completing the basic impedance matching of a single port, the goal of the second-stage optimization is to further minimize the maximum phase shift within the working frequency band and reduce the overall modulus of the reflection coefficient through full-station collaborative optimization.

[0137] In a possible implementation, the second stage achieves global optimal matching by comprehensively considering the coupling effect between ports and the dynamic priority weight.

[0138] Specifically, the objective function of the full-station collaborative optimization is defined as:

[0139] ;

[0140] where, is the impedance vector of all ports, is the frequency phase shift, is the reflection coefficient weight coefficient, is the working frequency band range of the substation, is the th port frequency modulus of the reflection coefficient.

[0141] The minimization of the objective function of the full-station collaborative optimization means achieving the minimum phase shift and the lowest reflection loss within the full frequency band. To solve this optimization problem, an iterative optimization algorithm is usually adopted. Exemplarily, the Newton-Raphson method or the genetic algorithm is used.

[0142] Finally, after the second-stage optimization is completed, the maximum phase shift within the working frequency band of the entire station should be controlled within the allowable range, and the modulus values of the reflection coefficients of each port are further reduced to ensure high-precision matching during the phase comparison process.

[0143] In the intelligent substation phase comparison method, S300 executes the phase comparison strategy, constructs a decision input set based on the optimized phase scattering model; designs an adaptive rule engine based on fuzzy logic and interface selection ideas, dynamically outputs the phase comparison working condition level, and automatically selects the phase comparison strategy accordingly.

[0144] Please refer to Figure 4 , which shows a flowchart of an exemplary intelligent substation phase comparison method S300 of the present application, and its content includes:

[0145] S310: Construct a decision input set based on the optimized phase scattering model.

[0146] The role of the decision input set is to convert complex phase matching information into executable decision parameters, providing a data basis for subsequent working condition level evaluation and strategy selection.

[0147] Extract parameters from the whole-station collaborative optimization model, including the phase shift amount, modulus value of the reflection coefficient, and impedance matching state of each port. Normalize the data to eliminate the influence of different dimensions. Exemplarily, convert the phase shift amount to a relative error percentage, and map the modulus value of the reflection coefficient to the interval [0,1]. At the same time, extract the dynamic priority of the basic factor output by the dynamic priority evaluation model in S100, and synchronously construct a decision input set containing three layers of data.

[0148] In a possible implementation manner, the constructed decision input set containing three layers of data includes:

[0149] Basic layer: Physical parameters directly from the whole-station collaborative optimization model, such as phase shift and reflection coefficient.

[0150] Status layer: Dynamic priority index of the basic factor output by the dynamic priority evaluation model.

[0151] Strategy layer: Compensation coefficient and priority weight driven by historical errors, used to guide resource allocation.

[0152] S320: Design an adaptive rule engine.

[0153] The goal of the adaptive rule engine is to dynamically generate the phase comparison working condition level (excellent / medium / poor) based on the decision input set, and achieve strategy matching through fuzzy logic and interface selection ideas.

[0154] In a possible implementation manner, the design steps of the adaptive rule engine include:

[0155] S321: Construct a fuzzy logic rule base.

[0156] Define input variables, including phase offset, reflection coefficient, and basic factor dynamic priority index.

[0157] Divide the fuzzy sets, and define three fuzzy intervals of "high", "medium", and "low" for each variable.

[0158] Formulate fuzzy rules. Exemplarily, "if the phase matching degree is high and the environment is stable, then the working condition level is excellent"; "if the equipment vibration is abnormal and historical errors are frequent, then the working condition level is poor".

[0159] S322: Dynamically match the phase comparison interface type based on the fuzzy inference result.

[0160] Specifically, for excellent working conditions, enable the high-speed synchronous interface to support full-node parallel phase comparison. For medium working conditions, switch to the low-bandwidth interface and perform time-sharing polling according to the historical error ranking. For poor working conditions, activate the redundant interface and start independent phase comparison for the primary and backup channels.

[0161] S330: Dynamically output the phase comparison execution strategy based on the decision input set and the adaptive rule engine.

[0162] In a possible implementation, the output phase comparison execution strategy includes:

[0163] S331: Excellent strategy. Use the high-speed synchronous interface to simultaneously start the phase calibration of all electrical connection ports, and rely on high-precision timestamps to ensure operation consistency. Dynamically allocate computing resources through the load balancing algorithm, and prioritize ensuring the matching speed of high-priority nodes such as the main transformer and busbar.

[0164] S332: Medium strategy. Sort the nodes according to the severity of historical errors, and prioritize calibrating the nodes with high-frequency errors. Exemplarily, prioritize calibrating the circuit breakers that have experienced impedance mismatch multiple times. Based on the time slice rotation strategy, allocate limited bandwidth resources to the nodes ranked at the top of the list to ensure that critical nodes complete the matching first. Monitor the matching progress of each node in real time. If a certain node completes ahead of schedule, immediately release the computing resources it occupies for subsequent nodes to use.

[0165] S333: Poor strategy. Adopt dual-channel redundant phase comparison and verification, where the primary channel performs conventional phase comparison, and the backup channel performs independent verification using a different algorithm. Compare the phase matching results of the primary and backup channels. If the deviation exceeds the threshold, trigger a secondary optimization process and record the exception log. If a certain channel is continuously abnormal, automatically switch to the backup channel and notify the operation and maintenance personnel to troubleshoot hardware failures.

[0166] In the intelligent substation phase comparison method, S400 implements differential compensation control for basic factors based on real-time dynamic priority weights. Slow-varying compensation is used for low-priority factors to avoid system oscillations, and mutation suppression is used for high-priority factors to ensure phase comparison stability. Finally, the compensation value is dynamically injected into the phase comparison strategy execution link.

[0167] Please refer to Figure 5 , which shows the flowchart of an exemplary intelligent substation phase comparison method S400 of the present application, and its content includes:

[0168] S410: Real-time monitoring and multi-dimensional quality assessment.

[0169] By deploying a distributed sensor network and edge computing nodes, dynamic parameters such as phase matching degree, reflection coefficient, device vibration amplitude, temperature and humidity, and historical error trigger frequency of each electrical connection port in the substation are collected in real time. Using multi-modal data fusion technology, the physical layer signals and environmental layer data are aligned in time and space to construct a multi-dimensional data stream containing timestamps, spatial positions, and parameter correlations.

[0170] In a possible implementation, based on a preset phase comparison quality assessment matrix, a three-level quality index system is defined, including:

[0171] Basic layer indicators include: port reflection coefficient modulus, phase offset, impedance normalization deviation;

[0172] State layer indicators include: dynamic priority weight volatility, historical error recurrence probability, device vibration abnormality;

[0173] Efficiency layer indicators include: phase comparison strategy resource utilization rate, interface switching response delay, and convergence speed of the whole station collaborative optimization.

[0174] Through the sliding window algorithm, weighted analysis of real-time data is carried out, combined with the typical working condition characteristics in the offline historical database, to dynamically identify the quality deviation trend of the current phase comparison process and generate a multi-dimensional quality assessment report, providing a decision basis for subsequent parameter optimization.

[0175] S420: Factor weight dynamic update mechanism.

[0176] Based on the quality assessment results output by S410, the incremental fuzzy comprehensive evaluation method is used to online correct the factor weights of the dynamic priority evaluation model in S100.

[0177] In a possible implementation, the online correction step includes:

[0178] S421: Normalize the voltage phase difference, temperature and humidity, vibration amplitude, and historical error data collected in real time to eliminate the dimension difference.

[0179] S422: Adjust the fuzzy interval parameters of the triangular membership function in S120 according to the deviation distribution characteristics in the latest quality assessment report, so that the membership degree curve can better fit the non-linear response characteristics of the current working condition.

[0180] S423: Introduce a time decay factor (0 < <1), perform weighted fusion on the historical weight vector and the current quality deviation gradient, and use the gradient descent algorithm to update the fuzzy weights of each factor. The formula is:

[0181] ;

[0182] where, is the fuzzy weight vector of the th basic factor at the th moment, is the quality assessment index, is the eigenvalue of the th basic factor.

[0183] S424: Recalculate the priority index according to the updated weight vector, perform real-time priority sorting on the voltage phase difference factor, environmental temperature and humidity factor, equipment vibration factor and historical error factor, and drive the strategy switching of the adaptive rule engine in S300.

[0184] S430: Fuzzy model parameter optimization.

[0185] Aiming at the static characteristics of the fuzzy logic rule base in S320, design a parameter self-tuning mechanism.

[0186] In a possible implementation manner, the design steps of the parameter self-tuning mechanism include:

[0187] S431: According to the probability density distribution of input variables such as phase shift and reflection coefficient in S410, use the kernel density estimation method to dynamically divide the "high / medium / low" fuzzy interval, replacing the fixed threshold set manually in the traditional way;

[0188] S432: Assign a dynamic confidence coefficient to the preset fuzzy rules in S321, and this coefficient is jointly calculated by the historical rule trigger accuracy rate and the current quality assessment deviation rate;

[0189] S433: Introduce a genetic algorithm to optimize the shape parameters of the fuzzy membership function, so that the fuzzy inference result can better fit the non-linear coupling effect under complex working conditions;

[0190] S434: When multiple fuzzy rules are triggered simultaneously and the conclusions are contradictory, the D-S evidence theory is used to fuse multi-source evidence, calculate the belief degree distribution of each conclusion, and output the working condition level determination with the maximum comprehensive belief degree.

[0191] S440: Adaptive adjustment of the phase comparison strategy.

[0192] According to the dynamic working condition level output by S430 and the factor priority updated by S420, perform the following strategy adjustments:

[0193] For excellent working conditions, enable the 5G high-speed synchronization interface to achieve parallel distribution and result aggregation of the phase calibration tasks for all station ports;

[0194] For intermediate working conditions, switch to the LoRa low-power wide-area network interface and schedule the calibration tasks in levels according to the severity of historical errors;

[0195] For poor working conditions, activate the redundant optical fiber channel and start the independent phase comparison and result cross-verification of the primary and standby dual channels;

[0196] S450: Abnormal mode recognition and compensation strategy generation.

[0197] Construct an abnormal mode recognition engine based on deep learning to extract features and perform pattern clustering on the multi-source data collected by S410.

[0198] In a possible implementation manner, a long short-term memory network is used to capture the long-term dependence relationships of parameters such as voltage phase difference and equipment vibration to identify periodic mismatch patterns. The electrical coupling relationships between ports in the substation topology are modeled through a graph neural network to locate the cascading mismatch caused by adjacent node failures;

[0199] The identified abnormal modes are stored in a labeled manner, and the compensation schemes with high-frequency recurrence are solidified as new fuzzy rules in the adaptive rule engine in S300. Exemplary: "If the vibration amplitude of a certain circuit breaker continuously exceeds the standard and the reflection coefficient > 0.3, then start the dynamic tuning of the reactance element".

[0200] Through the coordinated operation of the above sub-steps, S400 realizes a complete closed-loop from data collection, quality assessment to strategy evolution, enabling the phase comparison system to have the capabilities of equipment aging compensation, adaptation to environmental mutations, and self-identification of new fault modes, and finally achieving the leap of substation intelligent phase comparison technology from "precise matching" to "autonomous evolution".

[0201] This application also provides a substation intelligent phase comparison method system, including:

[0202] A dynamic priority evaluation module, which uses the voltage phase difference factor, the ambient temperature and humidity factor, the device vibration factor, and the historical error factor as basic factors, and judges the dynamic priority of the basic factors by constructing a dynamic priority evaluation model;

[0203] A multi-port modeling and optimization module, which, based on the substation topology, abstracts the signal transmission and reflection characteristics of each electrical connection port into a multi-port phase scattering model; analyzes the port phase scattering characteristics using the Smith chart, identifies impedance mismatch points, and implements two-stage optimization on the multi-port phase scattering model;

[0204] A phase verification strategy execution module, which executes the phase verification strategy, constructs a decision input set based on the optimized phase scattering model; designs an adaptive rule engine based on fuzzy logic and interface selection ideas, dynamically outputs the phase verification working condition level, and automatically selects the phase verification strategy accordingly;

[0205] A quality evaluation and optimization module, which conducts multi-dimensional quality evaluation by real-time monitoring of the port phase matching degree and reflection coefficient, and dynamically updates the factor weights and optimizes the fuzzy model parameters through the evaluation results to form a self-evolving system.

[0206] The basic principles of the present application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are only examples and not limitations, and it cannot be considered that these advantages, benefits, effects, etc. are essential for each embodiment of the present application. In addition, the above-disclosed specific details are only for illustrative purposes and for ease of understanding, rather than limitations. The above details do not limit the present application to necessarily adopt the above specific details for implementation.

[0207] The block diagrams of the devices, apparatuses, equipment, and systems involved in the present application are only illustrative examples and do not intend to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open-ended words, meaning "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the word "and / or", and can be used interchangeably with it, unless the context clearly indicates otherwise. The word "such as" used here refers to the phrase "such as but not limited to", and can be used interchangeably with it.

[0208] It should also be noted that in the devices, equipment, and methods of the present application, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent solutions of the present application.

[0209] The above description of the disclosed aspects enables any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0210] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. Substation intelligent phase comparison method, characterized in that, Including: Taking the voltage phase difference factor, ambient temperature and humidity factor, equipment vibration factor, and historical error factor as basic factors, and judging the dynamic priority of the basic factors by constructing a dynamic priority evaluation model; Based on the substation topology structure, abstracting the signal transmission and reflection characteristics of each electrical connection port into a multi-port phase scattering model; analyzing the port phase scattering characteristics using the Smith chart, identifying impedance mismatch points, and implementing two-stage optimization on the multi-port phase scattering model; Executing the phase comparison strategy and constructing a decision input set based on the optimized phase scattering model; Based on the fuzzy logic and interface selection idea, designing an adaptive rule engine to dynamically output the phase comparison working condition level, and automatically select the phase comparison strategy accordingly; Conducting multi-dimensional quality evaluation by real-time monitoring of the port phase matching degree and reflection coefficient, and dynamically updating the factor weights and optimizing the fuzzy model parameters through the evaluation results to form a self-evolving system.

2. The intelligent phase comparison method for a substation according to claim 1, wherein The dynamic priority evaluation model uses the fuzzy comprehensive evaluation method to calculate the weights of each basic factor, including: Establishing a triangular membership function, Gaussian membership function or S-type membership function to map the characteristic values of each basic factor to the [0,1] interval; Define the fuzzy weight vector. Through the formula Calculate the priority index, where is the weight of the th factor, is the membership degree of the th factor, is the eigenvalue of the th factor; Based on the priority index, dynamically sorting the voltage phase difference factor, ambient temperature and humidity factor, equipment vibration factor, and historical error factor to drive the adjustment of the phase comparison strategy.

3. The intelligent phase discrimination method for a substation according to claim 1, wherein, Abstract the electrical equipment in the substation as a multi-port network and define the scattering matrix as × complex matrix, where the element represents the scattering parameter of port for port ; Calculate the scattering parameters through the formula where is the scattering parameter, is the reflected wave of port , is the incident wave of the reflected wave of port . Combining the factor priority output by the dynamic priority evaluation model to perform weighted optimization on the multi-port phase scattering model.

4. The intelligent phase verification method for a substation according to claim 1, wherein The two-stage optimization includes: The first-stage optimization: dynamically adjusting the impedance parameters of high-priority ports by using parallel or series reactance elements, adopting a fixed compensation strategy for low-priority ports, and calculating the compensation value based on the feedback of the historical error factor; Second-stage optimization: Optimize the objective function through the collaboration of the entire station to achieve global optimal matching, where is the impedance vector of all ports, is the frequency is the phase offset, is the reflection coefficient weight coefficient, is the operating frequency band range of the substation, is the th port frequency is the modulus of the reflection coefficient.

5. The intelligent phase comparison method for a substation according to claim 1, wherein The decision input set contains three layers of data: The basic layer: including the phase offset, the modulus of the reflection coefficient, and the impedance matching state; The state layer: including the dynamic priority index of the basic factors output by the dynamic priority evaluation model; The strategy layer: including the compensation coefficient and priority weight driven by historical errors; Among them, the data is mapped to a unified dimension interval after normalization processing.

6. The intelligent phase comparison method for a substation according to claim 1, wherein The adaptive rule engine includes: Constructing a fuzzy logic rule base, defining the input variables as phase offset, reflection coefficient, and dynamic priority index of basic factors, and dividing the "high / medium / low" fuzzy intervals; Based on the fuzzy inference result, dynamically matching the phase comparison interface type, enabling the high-speed synchronous interface for the excellent working condition, switching to the low-bandwidth interface for the medium working condition, and activating the redundant interface for the poor working condition; Outputting the phase comparison execution strategy, including the full-node parallel calibration of the excellent strategy, the historical error sorting polling of the medium strategy, and the dual-channel redundant verification of the poor strategy.

7. The intelligent phase comparison method for a substation according to claim 1, wherein, The multi-dimensional quality evaluation includes: Defining a three-level quality index system: the basic layer indicators include the modulus of the port reflection coefficient, phase offset, and impedance normalization deviation; the state layer indicators include the dynamic priority weight volatility, historical error recurrence probability, and equipment vibration abnormality; the efficiency layer indicators include the resource utilization rate of the phase comparison strategy, interface switching response delay, and the convergence speed of the whole station collaborative optimization; Generating a multi-dimensional quality evaluation report by performing weighted analysis on real-time data through the sliding window algorithm.

8. The intelligent phase comparison method for a substation according to claim 1, wherein The dynamic update mechanism of the factor weights includes: Normalize the voltage phase difference, temperature and humidity, vibration amplitude, and historical error data collected in real time; Adjust the fuzzy interval parameters of the triangular membership function , and , where , and are fuzzy interval parameters, representing the lower limit, optimal value, and upper limit of the factor eigenvalue, respectively; Introduce a time decay factor , where 0 < < 1, and update the weights through the formula , where is the fuzzy weight vector of the -th basic factor at the -th moment, is the quality evaluation index, and is the eigenvalue of the -th basic factor.

9. The intelligent phase comparison method for a substation according to claim 1, wherein The fuzzy model parameter optimization includes: Dynamically divide the fuzzy interval using the kernel density estimation method; Assign dynamic confidence coefficients to the fuzzy rules, jointly calculated from the historical rule trigger accuracy and the current quality assessment deviation rate; Introduce the genetic algorithm to optimize the shape parameters of the fuzzy membership function; When multiple fuzzy rules conflict, use the D-S evidence theory to fuse multi-source evidence and output the working condition level determination with the maximum comprehensive trust degree.

10. The intelligent phase verification method for a substation according to claim 1, characterized in that, The steps for abnormal pattern recognition and compensation strategy generation in the self-evolution system include: Use the long short-term memory network to capture the long-term dependencies of voltage phase difference and equipment vibration; Model the electrical coupling relationship in the substation topology through a graph neural network; Store the identified abnormal patterns in a labeled manner, and solidify the frequently recurring compensation schemes as new fuzzy rules.

11. The substation intelligent phase discrimination method according to claim 1 or 2, characterized in that: The voltage phase difference factor is calculated through ; where is the voltage phase difference factor, is the actually measured voltage phase difference, is the maximum allowable voltage phase difference of the system; The ambient temperature and humidity factor is calculated through ; where is the ambient temperature and humidity factor, and are the weights of temperature and humidity respectively, and are the functions of the influence of temperature and humidity on the device performance respectively; The vibration factor of the device is calculated through ; where is the vibration factor of the device, and are the weights of the vibration frequency and amplitude respectively, and are the functions of the vibration frequency and amplitude on the device performance respectively; The historical error factor is calculated according to ; where is the historical error factor, is the weight of the -th historical error, is the severity of the -th historical error.

12. The intelligent phase verification method for a substation according to claim 1, characterized in that The phase discrimination execution strategy includes: Superior strategy: Rely on high-precision timestamps to achieve full-port parallel calibration, and prioritize ensuring the matching speed of the main transformer and bus through the load balancing algorithm; Intermediate strategy: Allocate limited bandwidth resources based on the time slice rotation strategy, and monitor the node progress in real time and dynamically release computing resources; Inferior strategy: The main and backup channels perform independent phase discrimination, trigger the secondary optimization process when the comparison result deviation occurs, automatically switch to the backup channel in case of abnormality, and notify the operation and maintenance personnel.

13. A system using the substation intelligent phase comparison method according to any one of claims 1-12, characterized in that, It includes: A dynamic priority evaluation module, which uses the voltage phase difference factor, ambient temperature and humidity factor, equipment vibration factor, and historical error factor as basic factors, and judges the dynamic priority of the basic factors by constructing a dynamic priority evaluation model; A multi-port modeling and optimization module, which abstracts the signal transmission and reflection characteristics of each electrical connection port into a multi-port phase scattering model based on the substation topology; uses the Smith chart to analyze the port phase scattering characteristics, identify the impedance mismatch points, and implement two-stage optimization on the multi-port phase scattering model; A phase discrimination strategy execution module, which is used to execute the phase discrimination strategy, construct a decision input set based on the optimized phase scattering model; design an adaptive rule engine based on fuzzy logic and interface selection ideas, dynamically output the phase discrimination working condition level, and automatically select the phase discrimination strategy accordingly; A quality evaluation and optimization module, which is used to perform multi-dimensional quality evaluation by real-time monitoring of the port phase matching degree and reflection coefficient, dynamically update the factor weights and optimize the fuzzy model parameters through the evaluation results to form a self-evolution system.

Images