Substation primary and secondary loop vector state comprehensive detection method and system

Abstract

The invention discloses a transformer substation primary and secondary loop vector state comprehensive detection method and system, and relates to the technical field of transformer substation automation, and the method comprises the following components: S1, a vector state comprehensive detection method, S2, intelligent data fusion and analysis, S3, automatic verification and self-diagnosis, and S4, operation process simplification and modular design. According to the method, the current, voltage and frequency electrical signals in the loop are collected in real time and converted into the vector form for comprehensive analysis, the amplitude of the signals is considered, the phase parameters of the signals are deeply analyzed, and therefore whether the phase, polarity and wiring of the loop are correct or not can be more accurately judged, and in addition, the accuracy of the loop is improved. The system further integrates multi-source data such as data, historical data and standard model data collected by a sensor in real time, real-time fusion and analysis are conducted through an intelligent algorithm, and the accuracy and timeliness of a detection result are further improved through comparison and correlation analysis of the multiple data sources.

Description

Technical Field

[0001] The present invention relates to the technical field of substation automation, and specifically provides a comprehensive detection method and system for the vector state of the primary and secondary circuits of a substation. Background Art

[0002] With the continuous development of the power system, as a key component of the power system, the safety and stability of the operating state of the substation are crucial for the reliable operation of the entire power grid. The primary and secondary circuits of the substation, as the core part of the substation, their state detection has always been the focus of power system maintenance. Traditional detection methods often rely on manual inspections and regular tests, which are not only time-consuming and laborious, but also difficult to reflect the state changes of the circuit in real time.

[0003] There are many disadvantages in the traditional state detection technology for the primary and secondary circuits of substations. First of all, the methods of manual inspection and regular test are inefficient and it is difficult to detect and handle potential problems in the circuit in a timely manner. Secondly, traditional methods often focus on the measurement and analysis of single electrical parameters, ignoring the phase relationship and polarity judgment between signals, resulting in limited accuracy of the detection results. In addition, traditional technologies are also easily interfered by human factors, such as operation errors and empirical judgments, which further affect the reliability of the detection results.

[0004] In summary, the traditional state detection technology for the primary and secondary circuits of substations has been difficult to meet the needs of modern power systems. In order to make up for the deficiencies of the existing technology and improve the detection efficiency and accuracy, therefore, it is particularly important to develop a comprehensive detection method and system for the vector state of the primary and secondary circuits of a substation. Summary of the Invention

[0005] The purpose of the present invention is to make up for the deficiencies of the existing technology and provide a comprehensive detection method and system for the vector state of the primary and secondary circuits of a substation. It can collect current, voltage, and frequency electrical signals in the circuit in real time, convert them into vector form for comprehensive analysis, not only considering the amplitude of the signals, but also deeply analyzing the phase parameters of the signals, so as to be able to more accurately judge whether the phase, polarity, and wiring of the circuit are correct. At the same time, the system also integrates multi-source data and uses intelligent algorithms for real-time fusion and analysis, further improving the accuracy and timeliness of the detection results.

[0006] In order to solve the above technical problems, the present invention provides the following technical solution: A comprehensive detection method for the vector state of the primary and secondary circuits of a substation. The specific steps of this method are as follows:

[0007] S1. Vector state comprehensive detection method: Deploy various sensors for current, voltage, and frequency in the primary and secondary circuits of a substation to collect electrical signals of current, voltage, and frequency in the circuit in real time. Convert the collected current and voltage signals into vector form, and combine vector analysis methods to comprehensively analyze the phase and amplitude parameters of each signal vector to determine whether the phase, polarity, and wiring of the circuit are correct;

[0008] S2. Intelligent data fusion and analysis: Integrate multi-source data such as data collected by sensors in real time, historical data, and standard model data, and use intelligent algorithms to perform real-time fusion and analysis on these data. By comparing and correlating information from different data sources, improve the accuracy and timeliness of the detection results, and eliminate human factor interference to the greatest extent;

[0009] S3. Automatic verification and self-diagnosis: Build an automatic verification and self-diagnosis module. During the detection process, automatically verify the detection results based on the standard model and historical data. Once it is found that there are wiring errors, polarity errors, or phase inconsistencies in the circuit that do not meet the standards, automatically mark the abnormal points, generate a detailed detection report, and send an alarm signal to provide clear handling suggestions for the staff;

[0010] S4. Simplify the operation process and modular design: Adopt wireless communication technology to enable on-site staff to remotely operate the detection system through a handheld device, simplify the equipment connection and operation process, and adopt a modular design concept. According to the detection requirements of different substations, flexibly combine each functional module so that it can be compatible with substations of different voltage levels and scales.

[0011] Furthermore, during the process of deploying various sensors in the primary and secondary circuits of the substation to collect electrical signals, a hierarchical distributed sensor deployment strategy is adopted. In the primary circuit, current sensors and voltage sensors are deployed for different electrical equipment. For the secondary circuit, corresponding sensors are deployed at the input and output ends of the protection device and the measurement and control device to comprehensively obtain electrical signals. The electrical signals collected by the sensors are preliminarily processed by the data acquisition unit, including signal amplification and filtering operations, to improve the signal quality. To improve the reliability of data acquisition, redundant acquisition technology is adopted, that is, for key electrical parameters, multiple sensors are deployed for acquisition. During the data transmission process, an encrypted communication protocol is adopted to ensure the security and integrity of the data, and the sampling frequency of the sensors is dynamically adjusted according to the actual operation conditions of the substation. During periods with frequent load changes, the sampling frequency is increased to capture more detailed electrical signal changes. Through the above measures of hierarchical distributed deployment, signal processing, redundant acquisition, encrypted communication, and dynamic sampling frequency adjustment, the electrical signals of the primary and secondary circuits of the substation can be collected more accurately and reliably.

[0012] Further, when converting the collected electrical signals into vector form, an adaptive vector conversion algorithm is adopted. Let the collected electrical signal be a time series x(t), and its discrete form be x[n], where n = 0, 1, …, N - 1, and N is the number of sampling points. First, windowing is performed on x[n] with a window function w[n] to obtain the windowed signal y[n] = x[n]w[n]. Then, fast Fourier transform is used to perform spectral analysis on y[n] to obtain the spectrum X[k], where k = 0, 1, …, N - 1. For the spectral component X[k0] corresponding to a specific frequency f0, its amplitude A and phase φ can be expressed as:

[0013] A = |X[k0]|

[0014] φ = ∠X[k0]

[0015] Next, an adaptive adjustment factor α is introduced, and its calculation formula is:

[0016]

[0017] where m is the statistical window length, is the mean value of the signals within the window, and the final vector representation is The source of the parameter m is determined through multiple experiments based on the fluctuation characteristics of substation electrical signals. Generally, its value ranges from 10 to 50. The adaptive adjustment factor α can adaptively adjust the amplitude of the vector according to the fluctuation of the signal, enabling the vector representation to more accurately reflect the actual state of the electrical signal. Through this adaptive vector conversion algorithm, the collected electrical signals can be more precisely converted into vector form, providing a more reliable basis for subsequent vector analysis.

[0018] Further, when combining the vector analysis method to judge whether the phase, polarity, and wiring of the circuit are correct, a comprehensive vector deviation evaluation algorithm is adopted. Let the standard vector be The vector actually detected is The amplitude deviation ΔA and phase deviation Δφ are defined as follows:

[0019] ΔA = |A act - A std |

[0020] Δφ = |φ act - φ std |

[0021] To comprehensively evaluate the deviation of the quantity, a comprehensive deviation index D is introduced, and its calculation formula is:

[0022]

[0023] Among them, w1 and w2 are the weight coefficients of the amplitude deviation and the phase deviation respectively, and w1 + w2 = 1. The sources of the weight coefficients w1 and w2 are determined by analyzing a large amount of historical fault data and using the regression analysis method in machine learning. Under different substation operation scenarios, the weight coefficients are dynamically adjusted according to the correlation degree between the fault type and the amplitude deviation and the phase deviation. In some cases, the phase deviation has a greater impact on the misoperation of the protection device, so the value of w2 is appropriately increased. When the comprehensive deviation index D exceeds the preset threshold T, it is determined that there is an abnormality in the circuit. Through this comprehensive vector deviation evaluation algorithm, it is possible to more comprehensively and accurately judge whether the phase, polarity, and wiring of the circuit are correct.

[0024] Furthermore, when fusing sensor data, historical data, and standard model data, a multi-source data dynamic fusion algorithm is adopted. Let the data collected by the sensor in real time be S, the historical data be H, and the standard model data be M. Data preprocessing is performed on S, H, and M, including data normalization and missing value processing. A time factor t is introduced to represent the timeliness of the data, and its value range is [0, 1]. The closer the time, the larger the value of t. For the sensor data S, its weight is w S = t. For the historical data H, its weight is w H = (1 - t) × β, where β is the credibility coefficient of the historical data, which is evaluated and determined according to the accuracy and integrity of the historical data. For the standard model data M, its weight is w M = 1 - w S - w H . The calculation formula for the fused data F is:

[0025] F = w S S + w H H + w M M

[0026] During the data fusion process, the time factor t and the credibility coefficient β are continuously updated to achieve the dynamic fusion of multi-source data. When the accuracy of the sensor data is relatively high, the value of w S is increased. When the credibility of the historical data is improved after verification, the value of β is increased. Through this multi-source data dynamic fusion algorithm, the real-time nature of the sensor data, the experience of the historical data, and the standardization of the standard model data can be fully utilized, improving the accuracy and reliability of the data fusion and providing better data support for subsequent intelligent analysis.

[0027] Furthermore, when the intelligent algorithm is used for real-time analysis, a fuzzy neural network anomaly diagnosis algorithm is adopted. This algorithm combines the advantages of fuzzy logic and neural networks, can more accurately identify the abnormal state in the circuit, constructs a fuzzy rule base, takes the amplitude and phase characteristics of the vector as input variables. When the amplitude deviation is large and the phase deviation is also large, it is determined as a serious anomaly. The input variables are fuzzified to obtain fuzzy inputs, and the fuzzy inputs are input into the neural network for training and learning. The neural network adopts a multi-layer perceptron structure, the number of nodes in its input layer is the number of input variables, the number of nodes in the hidden layer is adjusted according to the actual situation, and the number of nodes in the output layer is the number of anomaly types. The training process of the neural network adopts the backpropagation algorithm and introduces an adaptive learning rate adjustment strategy to improve the training efficiency and stability. Let the learning rate be η, and its calculation formula is:

[0028]

[0029] where η0 is the initial learning rate, E is the current training error, E max is the maximum allowable error. In the real-time analysis process, the fused data is input into the trained fuzzy neural network to obtain the anomaly diagnosis result. Through this fuzzy neural network anomaly diagnosis algorithm, multiple factors can be comprehensively considered, the abnormal state in the circuit can be more accurately identified, and the detection accuracy and reliability can be improved.

[0030] Furthermore, the automatic calibration and self-diagnosis module verifies the detection result based on the standard model and historical data, and adopts a dynamic threshold test algorithm. Let the eigenvalue of the detection result be x, and the eigenvalue range of the standard model is [x min , x max . The statistical eigenvalue of the historical data is the mean μ and the standard deviation σ. The threshold range of the standard model is dynamically adjusted according to the fluctuation of the historical data, and an adjustment factor γ is introduced. Its calculation formula is:

[0031]

[0032] The adjusted threshold range is [x min (1 - γ), x max (1 + γ)]. The eigenvalue x of the detection result is compared with the adjusted threshold range. If x is not within this range, it is determined that the detection result is abnormal. To avoid misjudgment, a confidence index C is introduced, and its calculation formula is:

[0033]

[0034] Where n1 is the number of historical data samples within the threshold range, and n2 is the number of historical data samples outside the threshold range. Only when the confidence level C is greater than the preset confidence level threshold C0 is the detection result considered abnormal. Through this dynamic threshold verification algorithm, the verification threshold can be dynamically adjusted according to the fluctuations of historical data, improving the accuracy and reliability of verification. At the same time, a confidence level indicator is introduced to avoid misjudgment.

[0035] Furthermore, when marking anomalies and generating reports, an anomaly classification marking and report generation algorithm is adopted. According to the severity of the anomaly and its impact on the operation of the substation, the anomalies are divided into four levels. Level 1 anomalies are minor anomalies with little impact on the operation of the substation. Level 2 anomalies are general anomalies that affect the normal operation of some equipment. Level 3 anomalies are severe anomalies that can cause misoperation of protection devices or equipment failures. Level 4 anomalies are critical anomalies that can trigger power grid accidents. Let the comprehensive evaluation index of the anomaly be R, and its calculation formula is:

[0036] R = w3D + w4C f

[0037] Where D is the comprehensive deviation index in the comprehensive vector deviation evaluation algorithm, and C f is the probability of a fault occurring, obtained through statistical analysis of historical fault data and real-time monitoring data. w3 and w4 are the weight coefficients of the comprehensive deviation index and the probability of a fault occurring respectively, and w3 + w4 = 1. The sources of the weight coefficients w3 and w4 are determined by using the analytic hierarchy process through the analysis of a large number of fault cases. According to the value of the comprehensive evaluation index R, the anomaly is marked as the corresponding level. When generating a report, the report content includes the basic information of the anomaly, the anomaly level, the analysis of the anomaly cause, and the treatment suggestions. The treatment suggestions are classified and provided according to the anomaly level and type. Through this anomaly classification marking and report generation algorithm, the anomaly situation can be displayed more clearly, providing more targeted treatment suggestions for operation and maintenance personnel.

[0038] Furthermore, during the entire detection process, a real-time data quality evaluation and correction algorithm is adopted to evaluate the quality of the real-time data collected by sensors to ensure the accuracy and reliability of the data. For data with poor quality, a correction method based on Kalman filtering is used for correction. Through the real-time data quality evaluation and correction algorithm, data with poor quality can be detected and corrected in a timely manner, improving the accuracy and reliability of the detection results.

[0039] On the other hand, a comprehensive vector state detection system for the primary and secondary circuits of a substation is characterized in that the system includes a vector state comprehensive detection module, an intelligent data fusion and analysis module, an automatic verification and self-diagnosis module, and a simplified operation process and modular design module:

[0040] The vector state comprehensive detection module: In the primary and secondary circuits of the substation, a variety of sensors for current, voltage, and frequency are arranged to collect the electrical signals of current, voltage, and frequency in the circuit in real time. The collected current and voltage signals are converted into vector forms, and combined with vector analysis methods, the phase and amplitude parameters of each signal vector are comprehensively analyzed to determine whether the phase, polarity, and wiring of the circuit are correct;

[0041] The intelligent data fusion and analysis module: Integrate multi-source data such as the data collected by sensors in real time, historical data, and standard model data, and use intelligent algorithms to perform real-time fusion and analysis on these data. By comparing and correlating the information of different data sources, the accuracy and timeliness of the detection results are improved, and the interference of human factors is excluded to the greatest extent;

[0042] The automatic calibration and self-diagnosis module: Construct an automatic calibration and self-diagnosis module. During the detection process, this module automatically verifies the detection results based on the standard model and historical data. Once it is found that there are wiring errors, polarity errors, or phase inconsistencies in the circuit that do not meet the standards, the abnormal points are automatically marked, and a detailed detection report is generated. At the same time, an alarm signal is sent to provide clear handling suggestions for the staff;

[0043] The simplified operation process and modular design module: Adopt wireless communication technology to enable on-site staff to remotely operate the detection system through handheld devices, simplify the equipment connection and operation process, and adopt the modular design concept. According to the detection requirements of different substations, each functional module can be flexibly combined to make it compatible with substations of different voltage levels and scales.

[0044] Compared with the prior art, the vector state comprehensive detection method and system for the primary and secondary circuits of a substation have the following beneficial effects:

[0045] First, the present invention collects the electrical signals of current, voltage, and frequency in the circuit in real time and converts them into vector forms for comprehensive analysis. This method not only considers the amplitude of the signals but also deeply analyzes the phase parameters of the signals, so as to more accurately determine whether the phase, polarity, and wiring of the circuit are correct. In addition, the system integrates multi-source data such as the data collected by sensors in real time, historical data, and standard model data, and uses intelligent algorithms for real-time fusion and analysis. This comparative and correlative analysis of multi-data sources further improves the accuracy and timeliness of the detection results. Through these measures, the present invention effectively reduces the interference of human factors, timely discovers potential problems in the circuit, and provides a strong guarantee for the safe operation of the substation.

[0046] II. Through wireless communication technology, on-site staff can conveniently perform remote operations on the detection system through a handheld device. This not only simplifies the device connection and operation process, but also reduces the operation difficulty and complexity. At the same time, the modular design concept enables the system to flexibly combine various functional modules according to the detection requirements of different substations, achieving compatibility with substations of different voltage levels and scales. This design not only improves the detection efficiency, but also makes the system more flexible, easy to expand and maintain. Therefore, the present invention has wide applicability and promotion value in practical applications.

[0047] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative work, other drawings can also be obtained based on these drawings.

[0049] Figure 1 It is a flow operation diagram of a comprehensive detection method for the vector state of the primary and secondary circuits of a substation;

[0050] Figure 2 It is a flow operation diagram of a comprehensive detection system for the vector state of the primary and secondary circuits of a substation. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] To further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention objective, the following, in combination with the drawings and preferred embodiments, details the specific embodiments, structures, features, and their effects of the present invention as follows.

[0052] Embodiment 1

[0053] This embodiment describes that in a newly built 110 kV substation, new transformer and switchgear equipment are about to be put into operation. To ensure the correct wiring of the primary and secondary circuits and normal vector state, a comprehensive detection is required.

[0054] According to the hierarchical distributed sensor layout strategy, current sensors are arranged on the high-voltage side of the transformer in the primary circuit, and voltage sensors are arranged on the bus side. Corresponding sensors are also installed at the input and output ends of the protection devices and measurement and control devices in the secondary circuit. Electrical signals are collected using redundant acquisition technology. After preliminary processing by the data acquisition unit, they are transformed into vector form using an adaptive vector conversion algorithm. For example, the discrete form of the voltage signal collected over a certain period is x[n], the number of sampling points N = 1024. After applying the window function w[n], we get y[n] = x[n]w[n]. Through the fast Fourier transform, we obtain the frequency spectrum X[k]. For the frequency spectrum component X[k0] corresponding to the rated frequency f0 = 50Hz, the amplitude A = |X[k0]|, and the phase Introduce an adaptive adjustment factor (assuming the statistical window length m = 128), and the final vector representation is Combined with the vector analysis method, the comprehensive vector deviation evaluation algorithm is used to judge the circuit conditions. Let the standard vector The actual detected vector The amplitude deviation ΔA = |A act - A std |, and the phase deviation The comprehensive deviation index (According to the scenario of new equipment commissioning, set the amplitude deviation weight w1 = 0.6 and the phase deviation weight w2 = 0.4). If D exceeds the set threshold, it is judged that there is a problem with the circuit.

[0055] Integrate the real-time acquisition data S of the sensors, historical data (since there is no historical operation data in this substation, assume the historical data H of a similar substation is used here), and standard model data M. First, preprocess the data, including normalization and missing value processing. Introduce the time factor t (for new equipment commissioning, the current time is the latest, t = 1), the weight w S of the sensor data = t = 1, the weight w H of the historical data = (1 - t)×β = 0 (assuming the credibility coefficient β of the historical data = 0.8), the weight w M of the standard model data = 1 - w S - w H = 0, and the fused data F = w S S + w H H + w M M = S. Use the fuzzy neural network anomaly diagnosis algorithm. Take the vector amplitude and phase characteristics as input variables. After fuzzy processing, input them into the neural network with a multi-layer perceptron structure. Let the initial learning rate η0 = 0.1 and the maximum allowable error E max = 0.01, the current training error E, and the learning rate After training, perform anomaly diagnosis on the fused data.

[0056] Using the automatic verification and self-diagnosis module, verify the test results based on the standard model and historical data (here, historical data of substations of the same type). Adopt the dynamic threshold test algorithm. Let the eigenvalue of the test result be x, and the eigenvalue range of the standard model be [x min , x max . The mean of the historical data is μ, the standard deviation is σ, and the adjustment factor . The adjusted threshold range is [x min (1 - γ), x max (1 + γ)]. Introduce the confidence index (n1 is the number of historical data samples within the threshold range, n2 is the number of samples outside the range). Preset the confidence threshold C0 = 0.9. If x is not within the adjusted threshold range and C > C0, then it is determined that the test result is abnormal. If there is an abnormality, adopt the abnormal grading marking and report generation algorithm. Let the comprehensive abnormal evaluation index R = w3D + w4C f (Assume that the weight of the comprehensive deviation index w3 = 0.7, the weight of the fault occurrence probability w4 = 0.3, and the fault occurrence probability C f = 0.2 is estimated according to experience). Determine the abnormal level according to the R value, and generate a report including abnormal information, level, cause analysis, and handling suggestions.

[0057] On-site staff remotely operate the detection system through a handheld wireless device, without complex on-site wiring connections. According to the scale of the 110 kV substation and the detection requirements, flexibly select functional modules such as voltage detection modules and current detection modules to achieve efficient detection.

[0058] Embodiment 2

[0059] This embodiment describes that a newly built 220 kV substation is about to be put into operation. To ensure the normal operation of the primary and secondary circuits, this detection method is used for comprehensive detection.

[0060] For the equipment of the new substation, sensors are arranged at key positions of each electrical equipment in the primary circuit and each device in the secondary circuit to collect the initial electrical signals before the new equipment is put into operation. After being converted into vector form, carefully analyze the vector parameters. For example, analyze the current and voltage vectors on the high and low voltage sides of the new transformer to determine whether the wiring meets the design requirements.

[0061] Since there is no historical data for the new substation, mainly fuse the real-time collected data and the standard model data. Use the multi-source data dynamic fusion algorithm, assign a higher weight to the real-time data for data fusion, and perform real-time analysis on the fused data through the fuzzy neural network abnormal diagnosis algorithm to establish a normal operation data model for the new equipment.

[0062] According to the standard model, the automatic verification and self-diagnosis module uses the dynamic threshold test algorithm to verify the detection results. During the load test of the new equipment, if the phase of the input current vector of a certain protection device is detected to be abnormal, the abnormal point is automatically marked, a report is generated and an alarm is issued, providing a processing direction for the debugging personnel.

[0063] On-site debugging personnel remotely operate the detection system through a handheld device to quickly complete various detection tasks. According to the scale and functional requirements of the new substation, they can flexibly select a module combination including functions such as high-precision data acquisition and comprehensive vector analysis to ensure that the detection system is adapted to the new substation.

[0064] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications within the scope of the technical solution of the present invention by using the disclosed technical content. These equivalent embodiments, as long as they do not depart from the technical solution content of the present invention, any simple modifications, equivalent changes and modifications made to the above embodiments according to the technical essence of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A comprehensive detection method for the vector state of the primary and secondary circuits of a substation, characterized in that, The specific steps of this method are as follows: S1. Vector state comprehensive detection method: In the primary and secondary circuits of a substation, various sensors such as current sensors, voltage sensors, and frequency sensors are arranged to collect electrical signals of current, voltage, and frequency in the circuit in real time. The collected current and voltage signals are converted into vector forms, and combined with vector analysis methods to comprehensively analyze the phase and amplitude parameters of each signal vector, and judge whether the phase, polarity, and wiring of the circuit are correct; S2. Intelligent data fusion and analysis: Integrate multi-source data such as data collected by sensors in real time, historical data, and standard model data, and use intelligent algorithms to perform real-time fusion and analysis on these data. By comparing and correlating information from different data sources, improve the accuracy and timeliness of the detection results, and exclude human factor interference to the greatest extent; S3. Automatic verification and self-diagnosis: Construct an automatic verification and self-diagnosis module. During the detection process, automatically verify the detection results based on the standard model and historical data. Once it is found that there are wiring errors, polarity errors, or phase inconsistencies in the circuit that do not meet the standards, automatically mark the abnormal points, generate a detailed detection report, and send an alarm signal at the same time; S4. Simplify the operation process and modular design: Adopt wireless communication technology to enable on-site staff to remotely operate the detection system through a handheld device, simplify the equipment connection and operation process, and adopt the modular design concept. According to the detection requirements of different substations, flexibly combine each functional module so that it can be compatible with substations of different voltage levels and different scales.

2. The integrated detection method for the vector state of the primary and secondary circuits of a substation according to claim 1, characterized in that During the process of arranging various sensors in the primary and secondary circuits of the substation to collect electrical signals, a hierarchical distributed sensor arrangement strategy is adopted. In the primary circuit, current sensors and voltage sensors are arranged for different electrical equipment. For the secondary circuit, corresponding sensors are arranged at the input and output ends of the protection device and the measurement and control device to comprehensively obtain electrical signals. The electrical signals collected by the sensors are preliminarily processed by the data acquisition unit. In order to improve the reliability of data acquisition, redundant acquisition technology is adopted, that is, for key electrical parameters, multiple sensors are arranged for acquisition. During the data transmission process, an encrypted communication protocol is adopted, and the sampling frequency of the sensors is dynamically adjusted according to the actual operation conditions of the substation. During periods with frequent load changes, the sampling frequency is increased to capture more detailed changes in electrical signals.

3. A comprehensive detection method for the vector state of the primary and secondary circuits of a substation according to claim 1, characterized in that, When converting the collected electrical signals into vector forms, an adaptive vector conversion algorithm is adopted. Let the collected electrical signal be the time series x(t), and its discrete form is xn, n = 0, 1, …, N - 1, where N is the number of sampling points. First, perform windowing processing on xn with the window function wn to obtain the windowed signal yn = xnwn. Then, perform spectral analysis on yn using the fast Fourier transform to obtain the spectrum Xk, k = 0, 1, …, N - 1. For the spectral component Xk0 corresponding to a specific frequency f0, its amplitude A and phase φ can be expressed as: A = |Xk0| φ = ∠Xk0 Next, introduce an adaptive adjustment factor α, and its calculation formula is: where m is the length of the statistical window, is the mean value of the signals within the window, and the final vector representation is The adaptive adjustment factor α can adaptively adjust the amplitude of the vector according to the fluctuation of the signal, so that the vector representation can more accurately reflect the actual state of the electrical signal.

4. A comprehensive detection method for the vector state of the primary and secondary circuits of a substation according to claim 1, characterized in that When using the combined vector analysis method to determine whether the phase, polarity, and wiring of the circuit are correct, a comprehensive vector deviation evaluation algorithm is adopted. Let the standard vector be The vector actually detected is The defined amplitude deviation ΔA and phase deviation Δφ are respectively:[ ΔA = |A act - A std | Δφ = |φ act - φ std | To comprehensively evaluate the deviation of the quantity, a comprehensive deviation index D is introduced, and its calculation formula is as follows: Among them, w1 and w2 are the weight coefficients of the amplitude deviation and the phase deviation respectively. Under different substation operation scenarios, the weight coefficients are dynamically adjusted according to the correlation degree between the fault type and the amplitude deviation and the phase deviation.

5. A comprehensive detection method for the vector state of the primary and secondary circuits of a substation according to claim 1, characterized in that, When fusing sensor data, historical data, and standard model data, a multi-source data dynamic fusion algorithm is adopted. Let the data collected by the sensor in real time be S, the historical data be H, and the standard model data be M. Data preprocessing is performed on S, H, and M, including data normalization and missing value processing. A time factor t is introduced to represent the timeliness of the data. The closer the time, the larger the t value. For the sensor data S, its weight is w S = t, for the historical data H, its weight is w H = (1 - t) × β, where β is the credibility coefficient of the historical data. For the standard model data M, its weight is w M = 1 - w S - w H , and the calculation formula for the fused data F is: F = w S S + w H H + w M M During the data fusion process, the time factor t and the credibility coefficient β are continuously updated to achieve the dynamic fusion of multi-source data.

6. A comprehensive detection method for the vector state of the primary and secondary circuits of a substation according to claim 1, characterized in that, When the intelligent algorithm mentioned above is used for real-time analysis, the fuzzy neural network anomaly diagnosis algorithm is adopted. This algorithm combines the advantages of fuzzy logic and neural networks and can more accurately identify the abnormal state in the circuit. A fuzzy rule base is constructed, and the amplitude and phase characteristics of the vector are used as input variables. The input variables are fuzzified to obtain fuzzy inputs, and the fuzzy inputs are input into the neural network for training and learning. The neural network adopts a multi-layer perceptron structure. The number of nodes in the input layer is the number of input variables, the number of nodes in the hidden layer is adjusted according to the actual situation, and the number of nodes in the output layer is the number of abnormal types. The training process of the neural network adopts the backpropagation algorithm and introduces an adaptive learning rate adjustment strategy. Let the learning rate be η, and its calculation formula is as follows: where η0 is the initial learning rate, E is the current training error, and E max is the maximum allowable error. During the real-time analysis process, the fused data is input into the trained fuzzy neural network to obtain the abnormal diagnosis result.

7. A comprehensive detection method for the vector state of the primary and secondary circuits of a substation according to claim 1, characterized in that, The automatic verification and self-diagnosis module verifies the detection results based on the standard model and historical data, and adopts a dynamic threshold test algorithm. Let the eigenvalue of the detection result be x, and the eigenvalue range of the standard model be x min , x max , the statistical eigenvalues of the historical data are the mean μ and the standard deviation σ. The threshold range of the standard model is dynamically adjusted according to the fluctuation of the historical data, and an adjustment factor γ is introduced. Its calculation formula is: The adjusted threshold range is x min 1 - γ, x max 1 + γ. Compare the eigenvalue x of the detection result with the adjusted threshold range. If x is not within this range, it is determined that there is an abnormality in the detection result. To avoid misjudgment, a confidence index C is introduced, and its calculation formula is as follows: Among them, n1 is the number of historical data samples within the threshold range, and n2 is the number of historical data samples outside the threshold range. When the confidence level C is greater than the preset confidence level threshold C0, the detection result is considered abnormal.

8. A comprehensive detection method for the vector state of the primary and secondary circuits of a substation according to claim 1, characterized in that, When marking the anomaly and generating a report, the anomaly grading marking and report generation algorithm is adopted. According to the severity of the anomaly and the impact on the substation operation, the anomalies are divided into four levels. Level I anomaly is a minor anomaly with little impact on the substation operation. Level II anomaly is a general anomaly that will affect the normal operation of some equipment. Level III anomaly is a serious anomaly that will cause the protection device to malfunction or equipment failure. Level IV anomaly is a critical anomaly that will trigger a power grid accident. Let the comprehensive evaluation index of the anomaly be R, and its calculation formula is as follows: R = w3D + w4C f where D is the comprehensive deviation index in the comprehensive vector deviation evaluation algorithm, and C f is the probability of the fault occurrence, w3 and w4 are the weight coefficients of the comprehensive deviation index and the probability of the fault occurrence respectively. According to the value of the comprehensive evaluation index R, the anomaly is marked with the corresponding level. When generating a report, the report content includes the basic information of the anomaly, the anomaly level, the analysis of the anomaly cause, and the handling suggestions. The handling suggestions are provided by classification according to the anomaly level and type.

9. A comprehensive detection method for the vector state of the primary and secondary circuits of a substation according to claim 1, characterized in that, During the entire detection process, the real-time data quality evaluation and correction algorithm is adopted to evaluate the quality of the real-time data collected by the sensor. For the data with poor quality, the correction method based on Kalman filtering is adopted for correction.

10. A comprehensive detection system for the vector state of the primary and secondary circuits of a substation, characterized in that, This system includes a vector state comprehensive detection module, an intelligent data fusion and analysis module, an automatic calibration and self-diagnosis module, and a simplified operation process and modular design module: The vector state comprehensive detection module: A variety of sensors such as current, voltage, and frequency are arranged in the primary and secondary circuits of the substation to collect the current, voltage, and frequency electrical signals in the circuit in real time. The collected current and voltage signals are converted into vector forms, and combined with vector analysis methods, the phase and amplitude parameters of each signal vector are comprehensively analyzed to judge whether the phase, polarity, and wiring of the circuit are correct; The intelligent data fusion and analysis module: Integrate multi-source data such as the data collected by the sensor in real time, historical data, and standard model data, and use intelligent algorithms to perform real-time fusion and analysis on these data. By comparing and correlating the information of different data sources, the accuracy and timeliness of the detection results are improved, and the interference of human factors is excluded to the greatest extent; The automatic verification and self-diagnosis module: An automatic verification and self-diagnosis module is constructed. During the detection process, this module automatically verifies the detection results based on the standard model and historical data. Once it is found that there are wiring errors, polarity errors, or phase inconsistencies in the circuit that do not meet the standards, it automatically marks the abnormal points, generates a detailed detection report, and at the same time issues an alarm signal, providing clear handling suggestions for the staff. The simplified operation process and modular design module: Wireless communication technology is adopted to enable on-site staff to remotely operate the detection system through a handheld device, simplifying the equipment connection and operation process. The modular design concept is adopted, and according to the detection requirements of different substations, each functional module can be flexibly combined to make it compatible with substations of different voltage levels and different scales.

Images