An automatic judgment and maintenance system for transformer high-voltage side phase loss fault

Abstract

The present invention provides a system for automatically determining and repairing a phase loss fault on the high-voltage side of a transformer, and relates to the field of data processing technology. The system is used to: calculate the effective value and phase difference of each phase voltage based on a preprocessed low-voltage side voltage data set, calculate the amplitude relationship coefficient, standard deviation, and phase characteristics of the three-phase voltage, obtain a phase loss characteristic data set, calculate the low-voltage side characteristic deviation factor based on the phase loss characteristic data and a preset phase loss model, and determine whether the three-phase voltage on the low-voltage side has a phase loss fault characteristic, calculate the amplitude ratio of each phase voltage, and compare it with the preset phase loss model to determine whether the amplitude ratio of the low-voltage side voltage meets the high-voltage side phase loss working condition, calculate the high-voltage side characteristic deviation factor, and compare it with the preset phase loss model to determine whether a phase loss fault exists on the high-voltage side, and obtain a phase loss determination result. The present invention can determine whether a phase loss fault occurs on the high-voltage side based on the low-voltage side voltage of the transformer.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an automatic determination and repair system for a transformer high-voltage side phase loss fault. Background Art

[0002] Transformers are core equipment in power systems and are widely used in transmission and distribution systems to achieve voltage conversion. During operation, a phase loss fault may occur on the high-voltage side of the transformer, resulting in voltage imbalance on the low-voltage side and affecting normal power supply. The current common phase loss detection method mainly relies on manual inspections, but this method is not only slow in detection speed, but also easily affected by the experience of operation and maintenance personnel, which may lead to unstable fault identification accuracy.

[0003] For example, in a certain region's 10kV distribution line, a Dyn11 connection group distribution transformer experienced phase loss due to a blown A-phase fuse on the high-voltage side. However, since the low-voltage side of the transformer was still able to output three-phase voltage, only the voltage amplitude changed. As a result, the on-site operation and maintenance personnel failed to immediately identify the phase loss fault when initially measuring the three-phase voltage on the low-voltage side, which may cause fault detection delays and affect power supply stability and safety. Summary of the Invention

[0004] The purpose of the present invention is to provide a system for automatically determining and repairing a transformer high-voltage side phase loss fault, aiming to solve the problems mentioned in the background technology.

[0005] In order to solve the above technical problems, the technical solutions of the present invention are as follows:

[0006] A system for automatically determining and repairing a transformer high-voltage side phase loss fault, the system comprising:

[0007] The voltage characteristic module is used to calculate the effective value and phase difference of each phase voltage based on the preprocessed low-voltage side voltage data set to obtain a voltage characteristic data set;

[0008] A phase loss feature module is used to calculate the amplitude relationship coefficient, standard deviation and phase characteristics of the three-phase voltage based on the voltage feature data set to obtain the phase loss feature data set;

[0009] The analysis module is used to calculate the low-voltage side characteristic deviation factor based on the phase loss characteristic data and the preset phase loss model, and determine whether the three-phase voltage on the low-voltage side has phase loss fault characteristics, thereby obtaining a low-voltage side phase loss abnormal signal;

[0010] A preliminary determination module is used to calculate the amplitude ratio of each phase voltage based on the low-voltage side phase loss abnormal signal, and compare it with the preset phase loss model to determine whether the amplitude ratio of the low-voltage side voltage meets the high-voltage side phase loss condition, thereby obtaining preliminary phase loss determination data;

[0011] The phase loss determination module is used to calculate the high-voltage side characteristic deviation factor based on the preliminary phase loss determination data, and compare it with the preset phase loss model to determine whether there is a phase loss fault on the high-voltage side and obtain the phase loss determination result;

[0012] The alarm module is used to determine the fault impact range based on the phase loss judgment result and the distribution network topology information, generate a phase loss fault signal, and send it to the maintenance end.

[0013] Furthermore, the voltage characteristic module includes:

[0014] The effective value calculation unit is used to perform a square operation on the instantaneous voltage data collected on the low-voltage side to obtain square voltage data; accumulate the square voltage data within a preset time length, and then take the average thereof to obtain mean square voltage data; perform a square root operation on the mean square voltage data to obtain effective value data of the three-phase voltage on the low-voltage side;

[0015] The phase difference calculation unit is used to convert the effective value data to obtain the fundamental phase data of the three-phase voltage; the phase difference data is obtained by calculating the difference between the fundamental phase data of the three-phase voltage.

[0016] Furthermore, the phase loss feature module includes:

[0017] A voltage amplitude relationship calculation unit is used to determine the maximum value, minimum value, and intermediate value of the three-phase voltage on the low-voltage side based on the effective value data to obtain low-voltage side voltage sorting data; based on the low-voltage side voltage sorting data, determine the sum of the intermediate value of the three-phase voltage and the minimum value of the three-phase voltage, and compare it with the maximum value of the three-phase voltage to obtain the low-voltage side voltage amplitude relationship coefficient;

[0018] a standard deviation calculation unit, configured to calculate the mean of the effective values of the three-phase voltages based on the effective value data to obtain mean data, and to calculate the deviation between the effective value data and the mean data of the three-phase voltages based on the mean data to obtain deviation data; and to calculate the variance and standard deviation of the three-phase voltages based on the deviation data to obtain standard deviation data;

[0019] The phase characteristic calculation unit is used to obtain phase characteristic data by calculating the absolute deviation value of the phase difference data of the three-phase voltage.

[0020] Furthermore, the analysis module includes:

[0021] A low-voltage side reference data acquisition unit is used to extract the historical phase-loss low-voltage side voltage database of the corresponding transformer model and connection group according to a preset phase-loss model to obtain low-voltage side reference data;

[0022] A low-voltage side characteristic deviation factor calculation unit is used to fuse the low-voltage side voltage change rate, the historical low-voltage side voltage change rate, the low-voltage side voltage standard deviation, and the historical low-voltage side voltage standard deviation based on the phase loss characteristic data and the low-voltage side reference data to obtain the low-voltage side characteristic deviation factor;

[0023] The characteristic analysis unit is used to judge whether the three-phase voltage on the low-voltage side meets the phase loss fault characteristics based on the low-voltage side characteristic deviation factor and the preset low-voltage side deviation threshold. When the result is yes, a low-voltage side phase loss abnormal signal is generated.

[0024] Furthermore, the preliminary determination module includes:

[0025] The amplitude ratio calculation unit is used to obtain total voltage data by calculating the sum of the three-phase voltage effective value data; and according to the total voltage data, respectively calculate the amplitude ratio of the three-phase voltage effective value data to the total voltage data to obtain a three-phase amplitude ratio data set;

[0026] The preliminary phase loss judgment unit is used to extract the low-voltage side amplitude ratio corresponding to the high-voltage side phase loss condition according to the preset phase loss model to obtain the phase loss amplitude ratio data set; based on the three-phase amplitude ratio data set and the phase loss amplitude ratio data set, it is judged whether the three-phase voltage amplitude ratio meets the high-voltage side phase loss condition. When the result is yes, the preliminary phase loss judgment data is obtained.

[0027] Furthermore, the phase loss determination module includes:

[0028] High-voltage side data acquisition unit, used to collect the voltage, current and power of the current transformer high-voltage side to obtain a real-time high-voltage side data set;

[0029] The high-voltage side parameter data acquisition unit is used to extract the historical phase-loss high-voltage side database of the corresponding transformer model and connection group according to the preset phase-loss model to obtain high-voltage side reference data;

[0030] A high-voltage side characteristic deviation factor calculation unit, configured to calculate a high-voltage side characteristic deviation factor based on a real-time high-voltage side data set and high-voltage side reference data;

[0031] The phase loss determination unit is used to determine whether there is a phase loss fault on the high-voltage side based on the high-voltage side characteristic deviation factor and the preset high-voltage side deviation threshold. When the result is yes, a phase loss determination result is generated.

[0032] Furthermore, the preset phase loss model includes:

[0033] The historical phase-loss low-voltage side voltage database is used to store the effective value, amplitude ratio and phase difference of the three-phase voltage on the low-voltage side during historical phase-loss faults;

[0034] The historical phase loss high-voltage side database is used to store the voltage RMS, current, and power values before and after the phase loss on the high-voltage side during historical phase loss faults;

[0035] The historical phase loss condition classification database is used to store historical phase loss fault types and their data.

[0036] Furthermore, the historical phase-missing low-voltage side voltage database includes:

[0037] Collect the three-phase voltage on the low-voltage side when the transformer phase failure occurs, and obtain the historical low-voltage side voltage RMS data;

[0038] According to the historical RMS voltage data of the low-voltage side, the amplitude ratio of the three-phase voltage on the low-voltage side when the phase failure occurs is calculated to obtain the historical phase failure voltage ratio range;

[0039] According to the historical RMS voltage data of the low-voltage side, the phase difference of the three-phase voltage on the low-voltage side when the phase loss fault occurs is calculated to obtain the historical phase loss phase angle variation range.

[0040] Furthermore, the historical phase-loss high-voltage side database includes:

[0041] Collect the effective value of the high-voltage side voltage before and after the phase failure occurs to obtain the historical effective value data of the high-voltage side voltage;

[0042] Collect the high-voltage side current values before and after the phase failure occurs to obtain historical high-voltage side current value data;

[0043] The high-voltage side power values before and after the phase failure occur are collected to obtain historical high-voltage side power value data.

[0044] Furthermore, the historical phase-loss condition classification database includes:

[0045] According to the transformer model and connection group, the historical phase-loss low-voltage side voltage data and the historical phase-loss high-voltage side voltage data are classified to obtain the connection type data;

[0046] According to the different phase loss faults on the high-voltage side, the historical phase loss low-voltage side voltage data and the historical phase loss high-voltage side voltage data are classified to obtain the phase loss type data;

[0047] According to the load conditions, the historical phase-loss low-voltage side voltage data and the historical phase-loss high-voltage side voltage data are classified to obtain the load type data.

[0048] The above solution of the present invention includes at least the following beneficial effects:

[0049] The present invention extracts the effective value of the voltage and phase difference of each phase from the low-voltage side voltage data set, obtains the voltage feature data set by calculation, and can convert the original voltage data into structured feature information. It not only realizes the accurate restoration of the three-phase voltage amplitude and phase state, but also improves the system's response capability to the change of the low-voltage side voltage state. When a phase loss occurs on the high-voltage side of the transformer, although the low-voltage side may still maintain three-phase output, the slight difference between the voltage amplitude and phase will be captured in time through this function to form a measurable feature indicator, thereby providing an accurate and quantifiable feature basis for subsequent modules. In particular, when it is impossible to intuitively judge whether the low-voltage side is abnormal in traditional methods, the pre-emptive and accuracy of phase loss detection are significantly improved through structured analysis, which provides a basic basis for establishing a multi-level judgment mechanism and effectively reduces the misjudgment and missed judgment caused by inaccurate manual experience.

[0050] The present invention constructs a phase-loss feature data set by calculating the amplitude relationship coefficient, standard deviation and phase characteristics of the three-phase voltage on the low-voltage side. By introducing the calculation of the amplitude relationship coefficient, the imbalance characteristics in the voltage can be effectively judged. Combined with the comprehensive judgment of the standard deviation and phase characteristics, a multi-dimensional abnormality judgment basis is constructed, which makes the system have a strong anti-interference ability. When there are interference factors such as instantaneous fluctuations and load changes in the power grid, conventional disturbances and structural abnormalities can be distinguished by statistical methods, thereby improving the robustness of phase-loss anomaly identification. In addition, the generated feature data has a certain normalization ability and model adaptability, and can adapt to the voltage status under different transformer models and operating environments, which significantly improves the overall generalization ability and deployment efficiency of the system.

[0051] The present invention calculates the low-voltage side characteristic deviation factor by introducing a fusion analysis method of historical data model and current phase-loss characteristic data, and judges whether the three-phase voltage on the low-voltage side has phase-loss characteristics based on this. It breaks through the limitation of traditional low-voltage side judgment relying on a single physical quantity, and realizes a fault identification mechanism based on multi-parameter fusion. The low-voltage side characteristic deviation factor comprehensively considers multiple dimensions such as voltage change rate, standard deviation, and historical data comparison to form a quantitative evaluation model for voltage anomalies, so that the system has the ability to identify global trend changes from local data, greatly improving the accuracy and timeliness of judgment. At the same time, an adaptive deviation threshold mechanism is introduced, which can adjust the judgment standard according to factors such as the connection group of the transformer and the historical operating status, thereby avoiding misidentification or missed detection.

[0052] The present invention calculates the amplitude ratio of the three-phase voltage and compares it with the typical characteristics in the preset phase loss model to determine whether the voltage amplitude ratio meets the typical operating conditions of phase loss on the high-voltage side, thereby forming preliminary phase loss judgment data. Through ratio processing, the interference of absolute voltage value fluctuations on fault identification is eliminated, so that the system can maintain a high recognition accuracy when the voltage level changes and the operating load differs greatly. The amplitude ratio construction method enables the judgment logic to have the ability of relative evaluation, and can identify phase loss characteristics through internal proportional relationships, which significantly improves the ability of phase loss identification to resist voltage fluctuations. Through the comparison mechanism with the preset phase loss model, the fault judgment result has a clear operating condition reference standard, avoiding the uncertainty caused by experience or threshold judgment, and effectively improving the overall accuracy and response speed of the system's multi-level judgment architecture.

[0053] The present invention uses preliminary phase loss judgment data to further collect real-time voltage, current and power data on the high-voltage side, and compares them with reference data in the historical database. Finally, by calculating the high-voltage side characteristic deviation factor, it determines whether there is a phase loss fault. The deep analysis capability of the high-voltage side operating status and the accuracy of fault location are improved. By introducing multi-dimensional real-time physical quantities as input data, it not only enhances the system's ability to understand the nature of the fault, but also improves the speed of identifying abnormal operating trends on the high-voltage side. The calculation mechanism of the characteristic deviation factor can obtain the degree of similarity between the current state and the phase loss model through quantitative analysis, ensuring that fault identification is not only real-time but also highly accurate. It can also call the corresponding phase loss model for comparison according to the differences in different wiring methods and transformer models, thereby realizing personalized judgment, enhancing the system's adaptability to complex power grid environments, and providing accurate information for subsequent maintenance.

[0054] Through modular structural design and multi-level data processing logic, the present invention constructs an automatic judgment and maintenance scheme that can judge the phase loss fault on the high-voltage side of the transformer in real time under the operating state. A bottom-up data flow and a top-down decision-making feedback mechanism are formed between the modules of the system, ensuring the logical consistency and data closed-loop of the entire process from data acquisition, feature extraction, model comparison to result output. The overall architecture not only improves the accuracy and real-time performance of the judgment results, but also solves the judgment lag problem caused by traditional reliance on manual inspection methods. Especially in different operating environments, different wiring methods and different load conditions, the system can automatically call the corresponding model for judgment based on actual parameters, enhancing the system's adaptability and generalization capabilities. At the same time, the system structure supports distributed deployment and can be widely applied to various application scenarios such as urban power grids, rural power grids and industrial parks. Through the automatic identification and positioning of voltage anomalies, the intelligence level of distribution network operation is significantly improved, providing strong support for fault warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a flowchart of a system for automatically determining and repairing a transformer high-voltage side phase loss fault provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0056] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0057] like Figure 1 As shown, an embodiment of the present invention provides a system for automatically determining and repairing a transformer high-voltage side phase loss fault, the system comprising:

[0058] The voltage characteristic module is used to calculate the effective value and phase difference of each phase voltage based on the preprocessed low-voltage side voltage data set to obtain a voltage characteristic data set;

[0059] A phase loss feature module is used to calculate the amplitude relationship coefficient, standard deviation and phase characteristics of the three-phase voltage based on the voltage feature data set to obtain the phase loss feature data set;

[0060] The analysis module is used to calculate the low-voltage side characteristic deviation factor based on the phase loss characteristic data and the preset phase loss model, and determine whether the three-phase voltage on the low-voltage side has phase loss fault characteristics, thereby obtaining a low-voltage side phase loss abnormal signal;

[0061] A preliminary determination module is used to calculate the amplitude ratio of each phase voltage based on the low-voltage side phase loss abnormal signal, and compare it with the preset phase loss model to determine whether the amplitude ratio of the low-voltage side voltage meets the high-voltage side phase loss condition, thereby obtaining preliminary phase loss determination data;

[0062] The phase loss determination module is used to calculate the high-voltage side characteristic deviation factor based on the preliminary phase loss determination data, and compare it with the preset phase loss model to determine whether there is a phase loss fault on the high-voltage side and obtain the phase loss determination result;

[0063] The alarm module is used to determine the fault impact range based on the phase loss judgment result and the distribution network topology information, generate a phase loss fault signal, and send it to the maintenance end.

[0064] In an embodiment of the present invention, the voltage characteristic module is used to calculate the effective value and phase difference of each phase voltage based on the preprocessed low-voltage side voltage data set to obtain a voltage characteristic data set, which can effectively capture the initial characteristics of voltage imbalance; the phase loss characteristic module is used to calculate the amplitude relationship coefficient, standard deviation and phase characteristics of the three-phase voltage based on the voltage characteristic data set to obtain a phase loss characteristic data set, thereby improving the system's ability to identify atypical phase loss faults; the analysis module is used to calculate the low-voltage side characteristic deviation factor based on the phase loss characteristic data and a preset phase loss model, and to determine whether the three-phase voltage on the low-voltage side has phase loss fault characteristics, thereby obtaining a low-voltage side phase loss abnormal signal and enhancing the ability to determine the changing trend of the voltage imbalance characteristics.

[0065] The preliminary judgment module is used to calculate the amplitude ratio of each phase voltage based on the abnormal phase loss signal on the low-voltage side, and compare it with the preset phase loss model to determine whether the amplitude ratio of the low-voltage side voltage meets the phase loss condition on the high-voltage side, and obtain preliminary phase loss judgment data. It can filter out the influence of absolute voltage fluctuation on the judgment accuracy, so as to more accurately identify the voltage imbalance structure caused by phase loss; the phase loss judgment module is used to calculate the high-voltage side characteristic deviation factor based on the preliminary phase loss judgment data, and compare it with the preset phase loss model to determine whether there is a phase loss fault on the high-voltage side, and obtain the phase loss judgment result, so as to avoid false alarm problems caused by low-voltage side load disturbance or harmonic pollution, and improve the accuracy of the system's final judgment result; the alarm module is used to determine the fault impact range based on the phase loss judgment result and combined with the distribution network topology information, generate a phase loss fault signal, and send it to the maintenance end, so as to realize rapid tracing and precise positioning of the fault, and significantly improve the fault response efficiency.

[0066] The alarm module is used to determine the fault impact range based on the phase loss determination result and the distribution network topology information, generate a phase loss fault signal, and send it to the maintenance end. Specifically, it includes:

[0067] First, the phase loss determination result from the phase loss determination module is received, and the number, real-time monitoring time and fault characteristic information of the transformer equipment with phase loss fault on the high-voltage side are obtained. Then, through the pre-stored distribution network topology database, the upstream and downstream electrical equipment information connected to the transformer with phase loss fault is retrieved, including adjacent transformer equipment, line branches and load node information, and the electrical connection relationship and spatial position distribution between the equipment are obtained, so as to preliminarily determine the scope of the power grid area that may be affected by the phase loss fault.

[0068] Next, the alarm module performs data fusion analysis on the topological structure information obtained above. By constructing a network topology model based on graph theory, it analyzes the possible extension path of the current phase failure in the distribution network, and uses the breadth-first search algorithm to traverse the distribution network topology map. Starting from the faulty device node, it expands layer by layer to determine the degree of impact of the fault signal on adjacent branches and load nodes. During the traversal process, the power supply dependency relationship of each node is marked, thereby accurately defining the scope of the fault area and the specific location of the users or power supply nodes that may be affected.

[0069] Then, the alarm module classifies the above traversal results into fault levels. Based on the preset fault severity assessment standards, it comprehensively considers multiple parameters such as the affected range, the importance of the load involved, and the number of affected users to determine the severity level of the phase loss fault. It matches the corresponding alarm mode and information content according to different levels and automatically generates a complete phase loss fault signal. The signal includes the fault transformer number, the time of fault occurrence, the fault level assessment, the specific range of the affected power supply area, and the name information of the important users or loads involved, so that subsequent maintenance work can be carried out accurately.

[0070] Finally, the alarm module automatically sends the complete phase loss fault signal through the communication network to the maintenance terminal, the dispatching control center, and the mobile maintenance terminal equipment of the operation and maintenance personnel in the corresponding area. At the same time, real-time recording and data archiving are carried out in the background of the monitoring system to achieve rapid notification and handling tracking of the fault.

[0071] In a preferred embodiment of the present invention, the voltage characteristic module includes:

[0072] The effective value calculation unit is used to perform a square operation on the instantaneous voltage data collected on the low-voltage side to obtain square voltage data; accumulate the square voltage data within a preset time length, and then take the average thereof to obtain mean square voltage data; perform a square root operation on the mean square voltage data to obtain effective value data of the three-phase voltage on the low-voltage side;

[0073] The phase difference calculation unit is used to convert the effective value data to obtain the fundamental phase data of the three-phase voltage; the phase difference data is obtained by calculating the difference between the fundamental phase data of the three-phase voltage.

[0074] In an embodiment of the present invention, the effective value calculation unit is used to perform a square operation on the instantaneous voltage data collected on the low-voltage side to obtain square voltage data; the square voltage data is accumulated within a preset time length, and then the mean is taken to obtain the mean square voltage data; the mean square voltage data is squared to obtain the effective value data of the three-phase voltage on the low-voltage side, which can directly characterize the stability and effectiveness of the actual output voltage on the low-voltage side of the transformer, so that the voltage characteristic data can accurately reflect the operating status of the transformer, providing an important quantitative basis for the subsequent accurate identification of phase loss faults; the phase difference calculation unit is used to convert the effective value data to obtain the fundamental phase data of the three-phase voltage; by calculating the difference between the fundamental phase data of the three-phase voltages, the phase difference data is obtained, thereby realizing the refined processing of the low-voltage side voltage characteristic data, and significantly improving the system's perception of abnormal voltage conditions on the low-voltage side, providing a solid and accurate feature foundation for the subsequent automatic detection of phase loss faults.

[0075] The phase difference calculation unit is used to convert the effective value data to obtain the fundamental phase data of the three-phase voltage; the phase difference data is obtained by calculating the difference between the fundamental phase data of the three-phase voltages, specifically including:

[0076] First, the three-phase voltage RMS data collected on the low-voltage side is preprocessed, and the collected time-domain RMS voltage data is Fourier transformed to extract the fundamental frequency component in each phase voltage signal, and then accurately calculate the phase value corresponding to the fundamental frequency. Specifically, the fast Fourier transform algorithm is used to accurately convert the original time-domain RMS signal into a frequency-domain signal, which significantly improves the accuracy of the fundamental phase data calculation and ensures that the subsequent analysis and judgment process based on phase information has higher stability and reliability.

[0077] After successfully obtaining the fundamental phase data of the three-phase voltage, the pairwise phase difference calculation is further performed, and the three-phase voltage fundamental phase data are combined and the difference calculation is performed respectively, that is, the difference between the fundamental phase of phase A voltage and the fundamental phase of phase B voltage, the difference between the fundamental phase of phase B voltage and the fundamental phase of phase C voltage, and the difference between the fundamental phase of phase C voltage and the fundamental phase of phase A voltage are calculated respectively to form three pairs of phase difference data. During the calculation, the phase value at the fundamental frequency is uniformly used as the calculation basis, which ensures the accuracy and uniformity of the phase difference data and avoids errors caused by inconsistent frequency components.

[0078] In a preferred embodiment of the present invention, the phase loss feature module includes:

[0079] A voltage amplitude relationship calculation unit is used to determine the maximum value, minimum value, and intermediate value of the three-phase voltage on the low-voltage side based on the effective value data to obtain low-voltage side voltage sorting data; based on the low-voltage side voltage sorting data, determine the sum of the intermediate value of the three-phase voltage and the minimum value of the three-phase voltage, and compare it with the maximum value of the three-phase voltage to obtain the low-voltage side voltage amplitude relationship coefficient;

[0080] a standard deviation calculation unit, configured to calculate the mean of the effective values of the three-phase voltages based on the effective value data to obtain mean data, and to calculate the deviation between the effective value data and the mean data of the three-phase voltages based on the mean data to obtain deviation data; and to calculate the variance and standard deviation of the three-phase voltages based on the deviation data to obtain standard deviation data;

[0081] The phase characteristic calculation unit is used to obtain phase characteristic data by calculating the absolute deviation value of the phase difference data of the three-phase voltage.

[0082] In an embodiment of the present invention, a voltage amplitude relationship calculation unit is used to determine the maximum value, minimum value and intermediate value of the three-phase voltage on the low-voltage side according to the effective value data, and obtain the low-voltage side voltage sorting data; according to the low-voltage side voltage sorting data, the sum of the intermediate value of the three-phase voltage and the minimum value of the three-phase voltage is determined, and compared with the maximum value of the three-phase voltage to obtain the low-voltage side voltage amplitude relationship coefficient, quantify the amplitude difference characteristics between the three-phase voltages on the low-voltage side, and accurately identify the amplitude abnormality trend caused by the phase failure; the standard deviation calculation unit is used to calculate the mean of the effective values of the three-phase voltages according to the effective value data to obtain the mean data. And based on the deviation between the effective value data and the mean data of the three-phase voltage, the deviation data is obtained; based on the deviation data, the variance and standard deviation of the three-phase voltage are calculated to obtain the standard deviation data, which eliminates the misjudgment interference caused by short-term fluctuations in the power grid and instantaneous changes in the load, and ensures that the identification of phase loss characteristics can be applied to complex operating conditions; the phase feature calculation unit is used to obtain the phase feature data by calculating the absolute deviation value of the phase difference data of the three-phase voltage, which can timely and sensitively capture the phase abnormality on the low-voltage side caused by the phase loss on the high-voltage side, and avoid the missed judgment or delayed judgment caused by the unclear voltage amplitude abnormality.

[0083] In a preferred embodiment of the present invention, the analysis module includes:

[0084] A low-voltage side reference data acquisition unit is used to extract the historical phase-loss low-voltage side voltage database of the corresponding transformer model and connection group according to a preset phase-loss model to obtain low-voltage side reference data;

[0085] A low-voltage side characteristic deviation factor calculation unit is used to fuse the low-voltage side voltage change rate, the historical low-voltage side voltage change rate, the low-voltage side voltage standard deviation, and the historical low-voltage side voltage standard deviation based on the phase loss characteristic data and the low-voltage side reference data to obtain the low-voltage side characteristic deviation factor;

[0086] The characteristic analysis unit is used to judge whether the three-phase voltage on the low-voltage side meets the phase loss fault characteristics based on the low-voltage side characteristic deviation factor and the preset low-voltage side deviation threshold. When the result is yes, a low-voltage side phase loss abnormal signal is generated.

[0087] In an embodiment of the present invention, a low-voltage side reference data acquisition unit is used to extract the historical phase-loss low-voltage side voltage database of the corresponding transformer model and wiring group according to a preset phase-loss model to obtain low-voltage side reference data, provide a historical reference standard for fault analysis, and eliminate the excessive reliance on experience in traditional manual diagnosis methods; a low-voltage side characteristic deviation factor calculation unit is used to fuse the low-voltage side voltage change rate, the historical low-voltage side voltage change rate, the low-voltage side voltage standard deviation and the historical low-voltage side voltage standard deviation according to the phase-loss characteristic data and the low-voltage side reference data to obtain a low-voltage side characteristic deviation factor, thereby realizing voltage anomaly detection. The quantitative characterization of the state ensures that fault analysis goes beyond qualitative judgment and truly achieves quantitative diagnosis, thereby improving the sensitivity to voltage anomalies and the accuracy of judgment, ensuring the quantitative distinction and accurate identification of fault states; the characteristic analysis unit is used to judge whether the three-phase voltage on the low-voltage side meets the phase loss fault characteristics based on the low-voltage side characteristic deviation factor and the preset low-voltage side deviation threshold. When the result is yes, a low-voltage side phase loss abnormality signal is generated, ensuring that the system's automatic identification of abnormal conditions has high accuracy, avoiding the subjective judgment errors that exist in traditional manual analysis, and significantly improving the degree of automation of system fault identification.

[0088] The low-voltage side characteristic deviation factor calculation unit is used to fuse the low-voltage side voltage change rate, the historical low-voltage side voltage change rate, the low-voltage side voltage standard deviation, and the historical low-voltage side voltage standard deviation based on the phase loss characteristic data and the low-voltage side reference data to obtain the low-voltage side characteristic deviation factor. Specifically, it includes:

[0089] The calculation formula of the low-pressure side characteristic deviation factor is:

[0090] ;

[0091] in, is the low-pressure side characteristic deviation factor, is the difference between the mean values of the low-voltage side voltage at adjacent time points, is the difference in historical low-voltage side voltages at adjacent time points, is the difference between adjacent time points, is the standard deviation of the low-voltage side voltage, is the historical low-voltage side voltage standard deviation, is the mean voltage on the low voltage side, is the historical low-voltage side voltage average, is the difference in standard deviation of the low-voltage side voltage at adjacent time points, is the difference in the standard deviation of the historical low-voltage side voltage at adjacent time points, and is the coefficient.

[0092] The above formula realizes the precise quantification and feature extraction of the difference between the current low-voltage side operating state of the transformer and the historical phase-loss fault state. First, the deviation comparison relationship between real-time and historical data is constructed by the difference between the mean low-voltage side voltage at adjacent time points and the difference between the historical low-voltage side voltage. Then, the fluctuation characteristic difference between real-time and historical data is constructed by using the difference between the standard deviation of the low-voltage side voltage at adjacent time points and the difference between the historical low-voltage side standard deviation.

[0093] In addition, the numerator of the formula uses an exponential function to perform further nonlinear amplification on the difference between the current voltage standard deviation and the historical voltage standard deviation, and performs weighted calculation based on the time interval. This exponential amplification allows the differences in voltage fluctuation characteristics to be magnified and presented more clearly, helping the system to quickly respond to small but critical voltage fluctuations and improving the sensitivity and recognition accuracy of early signs of phase loss faults.

[0094] The denominator of the formula divides the difference between the current and historical voltage mean values by the historical voltage mean value, and combines it with the voltage standard deviation difference for normalization and scaling. Adding a constant to the denominator can effectively prevent the abnormal amplification of the deviation factor caused by a denominator that is too small, making the calculation result more stable and reliable and avoiding the risk of false alarms.

[0095] Through the comprehensive calculation of the above formula, the low-voltage side characteristic deviation factor finally outputted can reflect the degree of deviation between the current operating voltage state and the historical phase loss fault state in a highly sensitive manner. The larger the value of this factor, the more significant the deviation between the current low-voltage side operating state and the typical phase loss fault characteristics, thereby clearly revealing that a phase loss fault may have occurred on the high-voltage side of the current transformer; otherwise, it means that the current state deviates little or not at all from the phase loss state, and can be judged as a normal or temporary fluctuation state.

[0096] Among them, the characteristic analysis unit is used to determine whether the three-phase voltage on the low-voltage side meets the phase loss fault characteristics based on the low-voltage side characteristic deviation factor and the preset low-voltage side deviation threshold. If the result is yes, a low-voltage side phase loss abnormality signal is generated, specifically including:

[0097] First, the low-voltage side deviation threshold is called from the preset database: according to the current transformer model, connection group and load characteristics, the threshold parameters that match the current actual operating status are selected from the pre-stored low-voltage side deviation threshold library. The called threshold is the statistical result of a large number of historical phase loss fault case analyses, which has certain representativeness and applicability, and can ensure the accuracy of the analysis unit's judgment results.

[0098] Then, the obtained low-voltage side characteristic deviation factor is numerically compared and analyzed with the low-voltage side deviation threshold; the system automatically performs numerical size comparison operations to determine in real time whether the actual value of the deviation factor exceeds the corresponding preset threshold; if the deviation factor exceeds the threshold, it indicates that the current low-voltage side voltage change trend or fluctuation range has exceeded the normal range, reflecting the typical voltage characteristics of phase loss abnormality; if it does not exceed the threshold, it means that the voltage state is in normal or allowable fluctuation range.

[0099] Finally, when the result of the comparison analysis exceeds the threshold, the feature analysis unit immediately and automatically generates a low-voltage side phase loss abnormality signal; and sends a clearly marked phase loss abnormality flag information to the subsequent preliminary judgment module. This information includes key data such as the currently detected deviation factor value, threshold information, transformer identification and timestamp.

[0100] In a preferred embodiment of the present invention, the preliminary determination module includes:

[0101] The amplitude ratio calculation unit is used to obtain total voltage data by calculating the sum of the three-phase voltage effective value data; and according to the total voltage data, respectively calculate the amplitude ratio of the three-phase voltage effective value data to the total voltage data to obtain a three-phase amplitude ratio data set;

[0102] The preliminary phase loss judgment unit is used to extract the low-voltage side amplitude ratio corresponding to the high-voltage side phase loss condition according to the preset phase loss model to obtain the phase loss amplitude ratio data set; based on the three-phase amplitude ratio data set and the phase loss amplitude ratio data set, it is judged whether the three-phase voltage amplitude ratio meets the high-voltage side phase loss condition. When the result is yes, the preliminary phase loss judgment data is obtained.

[0103] In an embodiment of the present invention, the amplitude ratio calculation unit is used to obtain total voltage data by calculating the sum of the three-phase voltage effective value data; based on the total voltage data, the amplitude ratios of the three-phase voltage effective value data and the total voltage data are respectively calculated to obtain a three-phase amplitude ratio data set, effectively extracting the relative proportional relationship of the three-phase voltage, so that the system can identify voltage asymmetry. This operation eliminates the interference caused by the absolute voltage value difference under different load states, and improves the sensitivity and recognition accuracy of real phase loss faults; the preliminary phase loss judgment unit is used to extract the low-voltage side amplitude ratio corresponding to the high-voltage side phase loss condition according to a preset phase loss model, and obtain a phase loss amplitude ratio data set; based on the three-phase amplitude ratio data set and the phase loss amplitude ratio data set, it is judged whether the three-phase voltage amplitude ratio meets the high-voltage side phase loss condition. When the result is yes, preliminary phase loss judgment data is obtained, and the voltage characteristic difference of the phase loss fault is accurately identified, effectively preventing the misjudgment of the voltage abnormality signal, and further improving the reliability and stability of phase loss fault identification.

[0104] Among them, the phase loss preliminary judgment unit is used to extract the low-voltage side amplitude ratio corresponding to the high-voltage side phase loss condition according to the preset phase loss model to obtain the phase loss amplitude ratio data set; based on the three-phase amplitude ratio data set and the phase loss amplitude ratio data set, it is judged whether the three-phase voltage amplitude ratio meets the high-voltage side phase loss condition. If the result is yes, the preliminary phase loss judgment data is obtained, which specifically includes:

[0105] First, based on the preset phase loss model, by accessing and querying the historical phase loss condition database, historical high-voltage side phase loss condition data consistent with the current transformer conditions are extracted from the database according to the transformer model, rated capacity and connection group. The historical condition data clearly includes the amplitude ratio range of the three-phase voltage corresponding to the low-voltage side when a phase loss occurs on the high-voltage side, thereby obtaining a phase loss amplitude ratio data set suitable for the current transformer model. According to the current transformer operating environment information, the amplitude ratio feature range corresponding to the current working condition is automatically retrieved and extracted from the database. This operation can realize the effective utilization of historical phase loss data and greatly improve the reliability and pertinence of data analysis.

[0106] Secondly, the phase loss amplitude ratio data set obtained in the previous step is compared and analyzed with the three-phase amplitude ratio data set obtained by real-time calculation. The specific operation method is: first, the amplitude ratio calculated in real time is taken for the three-phase voltages A, B, and C respectively, and then compared with the corresponding phase loss amplitude ratio range in the database to determine whether the real-time three-phase amplitude ratio falls within the typical phase loss amplitude ratio range recorded in the database. For example, the upper and lower limits of the amplitude ratio when the phase is lost are recorded in the database. and , the amplitude ratio calculated in real time is recorded as , then judge respectively Whether the inequality condition is satisfied By performing this operation on phases A, B, and C respectively, it is possible to quickly and clearly confirm whether the three-phase voltage amplitude ratio relationship is consistent with the historical phase-loss characteristic data, thereby realizing the preliminary identification of the phase-loss fault. When the real-time calculated amplitude ratio of at least one phase in the three-phase voltage amplitude ratio meets the phase-loss amplitude ratio range recorded in the preset phase-loss model, that is, when the real-time amplitude ratio data of the phase voltage falls within the characteristic data range, the preliminary phase-loss judgment unit confirms that the current voltage performance on the low-voltage side is consistent with the phase-loss operating condition characteristics on the high-voltage side.

[0107] In a preferred embodiment of the present invention, the phase loss determination module includes:

[0108] High-voltage side data acquisition unit, used to collect the voltage, current and power of the current transformer high-voltage side to obtain a real-time high-voltage side data set;

[0109] The high-voltage side parameter data acquisition unit is used to extract the historical phase-loss high-voltage side database of the corresponding transformer model and connection group according to the preset phase-loss model to obtain high-voltage side reference data;

[0110] A high-voltage side characteristic deviation factor calculation unit, configured to calculate a high-voltage side characteristic deviation factor based on a real-time high-voltage side data set and high-voltage side reference data;

[0111] The phase loss determination unit is used to determine whether there is a phase loss fault on the high-voltage side based on the high-voltage side characteristic deviation factor and the preset high-voltage side deviation threshold. When the result is yes, a phase loss determination result is generated.

[0112] In an embodiment of the present invention, a high-voltage side data acquisition unit is used to collect the voltage, current, and power of the current high-voltage side of the transformer to obtain a real-time high-voltage side data set. Through high-precision real-time data acquisition and processing, the timeliness and accuracy of the data in the system judgment process are guaranteed, and effective data guarantee is provided for subsequent feature analysis; a high-voltage side parameter data acquisition unit is used to extract the historical phase-loss high-voltage side database of the corresponding transformer model and connection group according to a preset phase-loss model to obtain high-voltage side reference data. By introducing historical data, the stability and robustness of the system judgment are improved, and the judgment deviation caused by transformer differences is effectively reduced; The high-voltage side characteristic deviation factor calculation unit is used to calculate the high-voltage side characteristic deviation factor based on the real-time high-voltage side data set and the high-voltage side reference data. The fusion calculation of multi-dimensional parameters significantly improves the accuracy of the high-voltage side phase loss judgment, and solves the technical defect that the single parameter judgment is easily affected by external interference; the phase loss judgment unit is used to judge whether there is a phase loss fault on the high-voltage side based on the high-voltage side characteristic deviation factor and the preset high-voltage side deviation threshold. When the result is yes, the phase loss judgment result is generated. Through threshold comparison and multi-dimensional cross-validation methods, the certainty of the high-voltage side phase loss judgment result is improved, and the false alarm or missed alarm problem caused by the abnormality of a single factor is avoided.

[0113] The high-voltage side characteristic deviation factor calculation unit is used to calculate the high-voltage side characteristic deviation factor based on the real-time high-voltage side data set and the high-voltage side reference data, specifically including:

[0114] The calculation formula of the high-voltage side characteristic deviation factor is:

[0115] ;

[0116] in, is the high-voltage side characteristic deviation factor, is the current voltage on the high-voltage side of the transformer, is the historical high-voltage side voltage, is the current on the high-voltage side of the transformer, is the current on the historical high-voltage side, is the active power on the high-voltage side of the current transformer, is the historical active power on the high voltage side, is the apparent power on the high voltage side of the current transformer, , is the coefficient.

[0117] Among them, the calculation formula performs difference, ratio and exponential operations between the current real-time operating parameters of the high-voltage side of the transformer and the historical operating data to obtain a quantitative indicator that can accurately reflect the degree of difference between the current operating status of the high-voltage side of the transformer and the historical normal operating status. The core function of the formula is to use precise mathematical calculations to reflect whether there is a phase loss fault on the high-voltage side with the combined difference value of multiple operating parameters, overcoming the limitations of single parameter judgment of traditional detection methods, so that abnormal operating conditions can be judged more precisely and intuitively, greatly improving the accuracy of fault judgment.

[0118] Specifically, the first term in the formula It reflects the relative magnitude of the difference between the current voltage and the historical voltage. By adjusting the sensitivity of the voltage difference, the formula can be adaptively adjusted for different transformer models and different voltage levels, ensuring accurate response to voltage anomalies under different operating conditions. The second term of the formula Reflect the change of current and control the weight of current difference, so as to flexibly optimize the contribution of current parameters to the judgment result in scenarios where the importance of current characteristics to phase loss judgment is different; the third item Reflecting power changes, by introducing the power change rate and adjusting the power characteristics, the robustness of the decision logic is significantly enhanced, the possibility of misjudgment caused by external disturbances is reduced, and the apparent power and exponential , effectively enhancing the nonlinear characterization capability of the abnormal degree of the current operating state, so that the deviation factor not only reflects the linear relationship between voltage, current and power, but also accurately captures the nonlinear change trend, thereby improving the sensitivity and robustness of fault identification.

[0119] Among them, according to the high-voltage side characteristic deviation factor and the preset high-voltage side deviation threshold, it is judged whether there is a phase loss fault on the high-voltage side. If the result is yes, a phase loss judgment result is generated, which specifically includes:

[0120] First, based on the calculation result of the high-voltage side characteristic deviation factor, the preset high-voltage side deviation threshold is retrieved from the system background preset model database. The preset high-voltage side deviation threshold is obtained after statistical analysis of the characteristic deviation factors corresponding to a large number of historical transformer high-voltage side phase loss faults. It is an important judgment basis for the system to determine the high-voltage side phase loss fault; then, the system compares the high-voltage side characteristic deviation factor with the preset high-voltage side deviation threshold item by item, and judges the difference between the current transformer high-voltage side operating state and the historical phase loss operating state in a quantitative manner, specifically including comparing the absolute value of the difference, the percentage of the difference or the comprehensive difference trend evaluation method, and through a strict data calculation process, it is clear whether the current state exceeds the allowable range of the normal operating state.

[0121] Secondly, if in the above comparative analysis, the high-voltage side characteristic deviation factor does not exceed the preset high-voltage side deviation threshold, the system determines that the current operating status of the high-voltage side of the transformer is normal, and continues to monitor the high-voltage side and perform periodic data comparison and analysis without generating additional actions; and when the calculated high-voltage side characteristic deviation factor exceeds the preset high-voltage side deviation threshold, the phase loss confirmation program set in the system is triggered, and the cross-validation mechanism is further initiated, including re-analysis of multiple sets of continuous periodic data and re-confirmation of the correlation between voltage, current, and power, to verify whether the characteristic deviation exists continuously and stably, thereby effectively avoiding misjudgment due to occasional instantaneous disturbances.

[0122] Finally, when the cross-validation mechanism confirms that the high-voltage side characteristic deviation factor has stably exceeded the preset threshold range for multiple consecutive cycles, and the multi-dimensional data verification results consistently meet the phase loss characteristic conditions, the system confirms that there is a phase loss fault on the high-voltage side of the transformer, and automatically calls the system's preset phase loss judgment result generation module, and accurately matches the current actual operating status of the transformer with the preset fault operating condition type in the model library to generate a clear and specific phase loss judgment result; the judgment result includes fault confirmation information, characteristic parameter information, fault severity information and preliminary positioning information, which is stored in the form of digital and structured data and further transmitted to the system's alarm module for sending alarm information to the maintenance terminal.

[0123] In a preferred embodiment of the present invention, the preset phase loss model includes:

[0124] The historical phase-loss low-voltage side voltage database is used to store the effective value, amplitude ratio and phase difference of the three-phase voltage on the low-voltage side during historical phase-loss faults;

[0125] The historical phase loss high-voltage side database is used to store the voltage RMS, current, and power values before and after the phase loss on the high-voltage side during historical phase loss faults;

[0126] The historical phase loss condition classification database is used to store historical phase loss fault types and their data.

[0127] In an embodiment of the present invention, a historical phase-loss low-voltage side voltage database is used to store the effective value, amplitude ratio and phase difference of the three-phase voltage on the low-voltage side during historical phase-loss faults, effectively establishing a reliable historical benchmark for low-voltage side voltage changes, and providing historical data support for the system to subsequently determine whether there is a phase-loss feature; a historical phase-loss high-voltage side database is used to store the effective value of the voltage, current value and power value before and after the phase loss on the high-voltage side during historical phase-loss faults. The system can accurately identify the significant changes in the electrical characteristics of the high-voltage side before and after the phase loss, thereby enhancing the identification accuracy of the phase-loss fault and improving the accuracy and reliability of fault location and diagnosis; a historical phase-loss operating condition classification database is used to store historical phase-loss fault types and their data, significantly improving the versatility and adaptability of the phase-loss fault detection and analysis system, and avoiding misjudgment or missed judgment due to equipment differences and different operating conditions.

[0128] In a preferred embodiment of the present invention, the historical phase-missing low-voltage side voltage database includes:

[0129] Collect the three-phase voltage on the low-voltage side when the transformer phase failure occurs, and obtain the historical low-voltage side voltage RMS data;

[0130] According to the historical RMS voltage data of the low-voltage side, the amplitude ratio of the three-phase voltage on the low-voltage side when the phase failure occurs is calculated to obtain the historical phase failure voltage ratio range;

[0131] According to the historical RMS voltage data of the low-voltage side, the phase difference of the three-phase voltage on the low-voltage side when the phase loss fault occurs is calculated to obtain the historical phase loss phase angle variation range.

[0132] In an embodiment of the present invention, the three-phase voltage on the low-voltage side is collected when a phase loss fault occurs on the transformer, and the historical effective value data of the low-voltage side voltage is obtained to ensure the reliability and accuracy of the subsequent voltage feature calculation, and provide strong data support for determining the phase loss fault; based on the historical effective value data of the low-voltage side voltage, the amplitude ratio of the three-phase voltage on the low-voltage side when the phase loss fault occurs is calculated to obtain the historical phase loss voltage ratio range, which can effectively distinguish between normal voltage fluctuations and amplitude anomalies caused by fault conditions, and provide a clear and referenceable voltage amplitude feature threshold for phase loss fault determination; based on the historical effective value data of the low-voltage side voltage, the phase difference of the three-phase voltage on the low-voltage side when the phase loss fault occurs is calculated to obtain the historical phase loss phase angle variation range, and clarify the phase loss feature to provide feature support in the phase dimension, which is particularly helpful to improve the sensitivity of the determination system to phase loss faults when the amplitude fluctuation is small.

[0133] The three-phase voltage on the low-voltage side when the transformer phase failure occurs is collected to obtain the historical low-voltage side voltage effective value data, including:

[0134] First, a high-precision voltage sensor is set at the output end of the low-voltage side of the transformer. The sensor can monitor and collect the instantaneous voltage data of the three phases A, B, and C on the low-voltage side of the transformer in real time. During implementation, the sampling frequency of the sensor is set to at least 128 sampling points per power frequency cycle to ensure the precision and real-time performance of data collection. The collected data is converted into a digital signal through the analog-to-digital conversion circuit, and then transmitted to the data processing terminal in real time. It is cached and stored according to the data unit of each power frequency cycle. In the system background, the voltage effective value calculation algorithm is used to process the cached instantaneous voltage data, that is, the instantaneous value of each phase voltage is squared in turn and the square average of all sampling points in each cycle is obtained, and then the square root processing is performed to accurately obtain the cycle effective value data of each phase voltage to form the historical low-voltage side voltage effective value data.

[0135] In a preferred embodiment of the present invention, the historical phase-loss high-voltage side database includes:

[0136] Collect the effective value of the high-voltage side voltage before and after the phase failure occurs to obtain the historical effective value data of the high-voltage side voltage;

[0137] Collect the high-voltage side current values before and after the phase failure occurs to obtain historical high-voltage side current value data;

[0138] The high-voltage side power values before and after the phase failure occur are collected to obtain historical high-voltage side power value data.

[0139] In an embodiment of the present invention, the effective value of the high-voltage side voltage before and after the phase loss fault occurs is collected to obtain historical high-voltage side effective value data, which can accurately record the specific change trend of the voltage value before and after the phase loss occurs, provide a quantitative comparison basis, and lay a data foundation for the subsequent establishment of a high-voltage side voltage characteristic model; the high-voltage side current value before and after the phase loss fault occurs is collected to obtain historical high-voltage side current value data. By comparing the high-voltage side current before and after the fault, the abnormal current fluctuation trend caused by the phase loss fault can be revealed, which serves as a basis for distinguishing the supplementary voltage information; the high-voltage side power value before and after the phase loss fault occurs is collected to obtain historical high-voltage side power value data. The power value is the embodiment of the electric energy output after the voltage and current are coupled, and can reflect the degree of influence of the phase loss on the overall load power supply capacity.

[0140] The effective values of the high-voltage side voltage before and after the phase failure are collected to obtain historical high-voltage side effective value data, including:

[0141] The system needs to be equipped with a voltage monitoring device with high-frequency sampling capability and installed at the high-voltage side incoming line end of the transformer to synchronously measure the three-phase voltages of A, B, and C. The sampling device is a voltage transformer, which records the instantaneous value of each phase voltage in real time with an acquisition period of no more than 20ms. The collected data is transmitted to the local monitoring platform or the remote master station system. To ensure data integrity, the system retains the voltage sampling values within the 10-second time period before and after the fault, and records them in the cache area according to the timestamp. The cached data is then calculated and processed for effective value: first, the instantaneous value of each phase voltage collected is squared, and then averaged in a sliding time window manner to obtain the mean square value of the time period; finally, the mean square value is squared to obtain the effective value sequence of each phase voltage. The system archives these voltage effective values in time series to form a historical high-voltage side voltage effective value data set.

[0142] The high-voltage side current values before and after the phase failure are collected to obtain historical high-voltage side current value data, including:

[0143] The system uses the current transformer installed on the high-voltage side of the transformer and cooperates with the electric energy meter to realize real-time acquisition of three-phase current signals. The system must ensure that the acquisition accuracy reaches 0.5S level or above, and the sampling frequency is not less than 1kHz to ensure the continuity and true reflection of the current fluctuation data before and after the fault. When the fault prediction module triggers the phase loss warning signal, the system automatically extracts the current sampling data 5 seconds before and after that moment. The system performs a sliding window operation on each phase current data. The steps include: first square processing, then averaging the data in each window, and then performing square root calculation to obtain the effective value of the current in the window. After the calculation is completed, the effective value of the current data is arranged in chronological order, and bound to the transformer number, phase information and fault event number, and uploaded to the high-voltage side historical current database.

[0144] The high-voltage side power values before and after the phase failure are collected to obtain historical high-voltage side power value data, including:

[0145] Based on the acquired voltage and current data, the system uses synchronous sampling to align the two in time domain, and completes the evaluation of power changes before and after the fault through the three-phase power calculation module. In each sampling cycle, the system performs the following operations: For each phase, calculate the active power , reactive power , apparent power ,in is the effective value of voltage and current at the corresponding moment, The phase difference between voltage and current is calculated from the time delay of synchronous sampling. The power of each phase is then summed to obtain the total three-phase power value. The system records power data for at least 10 sampling periods before and after the data is collected and stores the power change curve in the historical high-voltage side power value data. Each set of power data is associated with the transformer number, sampling period, and phase loss type label to facilitate subsequent retrieval by fault scenario classification.

[0146] In a preferred embodiment of the present invention, the historical phase-loss condition classification database includes:

[0147] According to the transformer model and connection group, the historical phase-loss low-voltage side voltage data and the historical phase-loss high-voltage side voltage data are classified to obtain the connection type data;

[0148] According to the different phase loss faults on the high-voltage side, the historical phase loss low-voltage side voltage data and the historical phase loss high-voltage side voltage data are classified to obtain the phase loss type data;

[0149] According to the load conditions, the historical phase-loss low-voltage side voltage data and the historical phase-loss high-voltage side voltage data are classified to obtain the load type data.

[0150] In an embodiment of the present invention, historical phase-loss low-voltage side voltage data and historical phase-loss high-voltage side voltage data are classified according to the transformer model and wiring group to obtain wiring type data. By introducing the wiring group dimension to classify the data, the subsequent model call process can select a matching phase-loss feature model based on the actual wiring method of the transformer, thereby improving the pertinence and accuracy of the analysis results; according to different phase loss faults on the high-voltage side, the historical phase-loss low-voltage side voltage data and historical phase-loss high-voltage side voltage data are classified to obtain phase-loss type data. Through the refined classification of phase loss types, the system can establish model templates for different phase loss situations, thereby improving the ability to recognize fault modes in the phase loss judgment process; according to the load condition, the historical phase-loss low-voltage side voltage data and historical phase-loss high-voltage side voltage data are classified to obtain load type data, which can enable the system to introduce a condition correction mechanism in the judgment process, thereby improving the ability of the model to adapt to complex operating environments.

[0151] Among them, according to the transformer model and connection group, the historical phase-loss low-voltage side voltage data and the historical phase-loss high-voltage side voltage data are classified to obtain the connection type data, which specifically includes:

[0152] First, access the transformer asset management system or operation ledger database through the data interface to extract the unique identification number, equipment model information and wiring group parameters of each transformer. The wiring groups usually include Yyn0, Dyn11, Yd11 and other types. The system establishes a wiring group index table and labels transformers with the same wiring group with a unified category label.

[0153] Subsequently, by associating queries with the phase loss fault record database, the voltage data corresponding to all phase loss faults are screened out, including the effective value and phase angle of the three-phase voltage on the low-voltage side, and the three-phase voltage, current and power parameters on the high-voltage side. The system binds each set of phase loss fault data with the connection group of the transformer to which it belongs, and classifies and stores them according to the connection method.

[0154] Among them, according to the different phase loss faults on the high-voltage side, the historical phase loss low-voltage side voltage data and the historical phase loss high-voltage side voltage data are classified to obtain the phase loss type data, which specifically includes:

[0155] The system first extracts the type of phase missing on the high-voltage side through the fault labeling information recorded in the phase loss event data table, including labels such as missing phase A, missing phase B, and missing phase C. If the original record does not show the missing phase labeled, the system will automatically analyze the effective value data of the high-voltage side voltage, automatically infer the missing phase based on the characteristic that the voltage of the missing phase drops to zero or is significantly lower, and supplement the missing label in a programmatic manner.

[0156] After completing phase identification, the system creates a corresponding data index file for each type of phase loss, and aggregates the low-voltage and high-voltage side voltage data, current data, and power data corresponding to the phase A loss event into one category; independent data sets are established for phase B loss and phase C loss respectively. For power systems with large data volumes, further subdivision can be performed, such as cross-labeling such as phase A loss - wiring group is Dyn11, load type is heavy load, etc.

[0157] The historical phase-loss low-voltage side voltage data and the historical phase-loss high-voltage side voltage data are classified according to the load conditions to obtain load type data, including:

[0158] The system first extracts the load information data corresponding to the phase loss event from the distribution automation system, including the three-phase current value, instantaneous power value, active power and reactive power components, and power factor. By analyzing the load-side electrical parameter curve during this time period, the system can calculate the average load level, load fluctuation degree and current imbalance factor during this period.

[0159] Based on these calculation results, the system divides load conditions into several categories, such as heavy load operation, light load operation, and fluctuating load operation. At the same time, the system also divides load conditions into resistive, inductive, capacitive, or composite load types based on load characteristics. The power factor range is used as the judgment basis. The power factor of inductive load is lower than 0.85, and that of resistive load is close to 1. After the classification is completed, the system archives the voltage, current, and power data collected from the phase loss events that occurred under each type of load condition, and associates the wiring group with the phase loss type label to build a multi-dimensional label classification data model.

[0160] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A transformer high-voltage side phase loss fault automatic determination and maintenance system, characterized by: The system comprises: The voltage characteristic module is used to calculate the effective value and phase difference of each phase voltage based on the preprocessed low-voltage side voltage data set to obtain a voltage characteristic data set; A phase loss feature module is used to calculate the amplitude relationship coefficient, standard deviation and phase characteristics of the three-phase voltage based on the voltage feature data set to obtain the phase loss feature data set; The analysis module is used to calculate the low-voltage side characteristic deviation factor based on the phase loss characteristic data and the preset phase loss model, and determine whether the three-phase voltage on the low-voltage side has phase loss fault characteristics, thereby obtaining a low-voltage side phase loss abnormal signal; A preliminary determination module is used to calculate the amplitude ratio of each phase voltage based on the low-voltage side phase loss abnormal signal, and compare it with the preset phase loss model to determine whether the amplitude ratio of the low-voltage side voltage meets the high-voltage side phase loss condition, thereby obtaining preliminary phase loss determination data; The phase loss determination module is used to calculate the high-voltage side characteristic deviation factor based on the preliminary phase loss determination data, and compare it with the preset phase loss model to determine whether there is a phase loss fault on the high-voltage side and obtain the phase loss determination result; The alarm module is used to determine the fault impact range based on the phase loss judgment result and the distribution network topology information, generate a phase loss fault signal, and send it to the maintenance end.

2. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 1 is characterized in that: The voltage characteristic module includes: The effective value calculation unit is used to perform a square operation on the instantaneous voltage data collected on the low-voltage side to obtain square voltage data; accumulate the square voltage data within a preset time length, and then take the average thereof to obtain mean square voltage data; perform a square root operation on the mean square voltage data to obtain effective value data of the three-phase voltage on the low-voltage side; The phase difference calculation unit is used to convert the effective value data to obtain the fundamental phase data of the three-phase voltage; the phase difference data is obtained by calculating the difference between the fundamental phase data of the three-phase voltage.

3. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 2 is characterized in that: The phase loss feature module includes: A voltage amplitude relationship calculation unit is used to determine the maximum value, minimum value, and intermediate value of the three-phase voltage on the low-voltage side based on the effective value data to obtain low-voltage side voltage sorting data; based on the low-voltage side voltage sorting data, determine the sum of the intermediate value of the three-phase voltage and the minimum value of the three-phase voltage, and compare it with the maximum value of the three-phase voltage to obtain the low-voltage side voltage amplitude relationship coefficient; a standard deviation calculation unit, configured to calculate the mean of the effective values of the three-phase voltages based on the effective value data to obtain mean data, and to calculate the deviation between the effective value data and the mean data of the three-phase voltages based on the mean data to obtain deviation data; and to calculate the variance and standard deviation of the three-phase voltages based on the deviation data to obtain standard deviation data; The phase characteristic calculation unit is used to obtain phase characteristic data by calculating the absolute deviation value of the phase difference data of the three-phase voltage.

4. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 3 is characterized in that: The analysis module includes: A low-voltage side reference data acquisition unit is used to extract the historical phase-loss low-voltage side voltage database of the corresponding transformer model and connection group according to a preset phase-loss model to obtain low-voltage side reference data; A low-voltage side characteristic deviation factor calculation unit is used to fuse the low-voltage side voltage change rate, the historical low-voltage side voltage change rate, the low-voltage side voltage standard deviation, and the historical low-voltage side voltage standard deviation based on the phase loss characteristic data and the low-voltage side reference data to obtain the low-voltage side characteristic deviation factor; The characteristic analysis unit is used to judge whether the three-phase voltage on the low-voltage side meets the phase loss fault characteristics based on the low-voltage side characteristic deviation factor and the preset low-voltage side deviation threshold. When the result is yes, a low-voltage side phase loss abnormal signal is generated.

5. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 4 is characterized in that: The preliminary determination module includes: The amplitude ratio calculation unit is used to obtain total voltage data by calculating the sum of the three-phase voltage effective value data; and according to the total voltage data, respectively calculate the amplitude ratio of the three-phase voltage effective value data to the total voltage data to obtain a three-phase amplitude ratio data set; The preliminary phase loss judgment unit is used to extract the low-voltage side amplitude ratio corresponding to the high-voltage side phase loss condition according to the preset phase loss model to obtain the phase loss amplitude ratio data set; based on the three-phase amplitude ratio data set and the phase loss amplitude ratio data set, it is judged whether the three-phase voltage amplitude ratio meets the high-voltage side phase loss condition. When the result is yes, the preliminary phase loss judgment data is obtained.

6. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 5 is characterized in that: The phase loss determination module includes: High-voltage side data acquisition unit, used to collect the voltage, current and power of the current transformer high-voltage side to obtain a real-time high-voltage side data set; The high-voltage side parameter data acquisition unit is used to extract the historical phase-loss high-voltage side database of the corresponding transformer model and connection group according to the preset phase-loss model to obtain high-voltage side reference data; A high-voltage side characteristic deviation factor calculation unit, configured to calculate a high-voltage side characteristic deviation factor based on a real-time high-voltage side data set and high-voltage side reference data; The phase loss determination unit is used to determine whether there is a phase loss fault on the high-voltage side based on the high-voltage side characteristic deviation factor and the preset high-voltage side deviation threshold. When the result is yes, a phase loss determination result is generated.

7. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 6 is characterized in that: The preset phase loss model includes: The historical phase-loss low-voltage side voltage database is used to store the effective value, amplitude ratio and phase difference of the three-phase voltage on the low-voltage side during historical phase-loss faults; The historical phase loss high-voltage side database is used to store the voltage RMS, current, and power values before and after the phase loss on the high-voltage side during historical phase loss faults; The historical phase loss condition classification database is used to store historical phase loss fault types and their data.

8. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 7 is characterized in that: The historical phase-missing low-voltage side voltage database includes: Collect the three-phase voltage on the low-voltage side when the transformer phase failure occurs, and obtain the historical low-voltage side voltage RMS data; According to the historical RMS voltage data of the low-voltage side, the amplitude ratio of the three-phase voltage on the low-voltage side when the phase failure occurs is calculated to obtain the historical phase failure voltage ratio range; According to the historical RMS voltage data of the low-voltage side, the phase difference of the three-phase voltage on the low-voltage side when the phase loss fault occurs is calculated to obtain the historical phase loss phase angle variation range.

9. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 8 is characterized in that: The historical phase-loss high-voltage side database includes: Collect the effective value of the high-voltage side voltage before and after the phase failure occurs to obtain the historical effective value data of the high-voltage side voltage; Collect the high-voltage side current values before and after the phase failure occurs to obtain historical high-voltage side current value data; The high-voltage side power values before and after the phase failure occur are collected to obtain historical high-voltage side power value data.

10. The automatic determination and maintenance system for a transformer high-voltage side phase loss fault according to claim 9 is characterized in that: The historical phase-loss condition classification database includes: According to the transformer model and connection group, the historical phase-loss low-voltage side voltage data and the historical phase-loss high-voltage side voltage data are classified to obtain the connection type data; According to the different phase loss faults on the high-voltage side, the historical phase loss low-voltage side voltage data and the historical phase loss high-voltage side voltage data are classified to obtain the phase loss type data; According to the load conditions, the historical phase-loss low-voltage side voltage data and the historical phase-loss high-voltage side voltage data are classified to obtain the load type data.

Images