Method for double feature identification of slight inter-turn short circuit of high voltage winding of micro temperature signal

Abstract

The present application relates to the field of power equipment fault detection, and specifically discloses a double-feature identification method for slight inter-turn short circuit of high-voltage winding with slight temperature signal, which uses a non-contact infrared sensing array to realize micro-temperature collection in a narrow space and synchronously acquires three-phase current signals; after interference elimination, micro-temperature difference calculation and micro-signal amplification, standard temperature features are extracted, and after pre-processing of the current signals, phase difference standard features are obtained; a temperature-phase difference coupling feature vector is constructed based on the temperature influence mechanism, the double-feature abnormal correlation degree is calculated, and fault identification is completed according to the grading threshold, and the reference feature library is updated in real time. The present application solves the problems of difficulty in temperature measurement in the narrow space of high-voltage winding, easy interference of micro-signal, and disconnection of temperature and electrical features, retains the innovation point of double-feature fusion, realizes early and accurate identification of implicit inter-turn short circuit fault, has low misjudgment and omission rate, and is suitable for online monitoring of high-voltage motor and generator stator winding, and has strong engineering practicability.

Description

Technical Field

[0001] This invention relates to the field of power equipment fault detection technology, specifically a dual-feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals. Background Technology

[0002] The high-voltage stator winding is a key component of core power equipment such as generators and high-voltage motors. Minor inter-turn short circuits are the most common latent insulation faults that occur in the early stages of operation. Specifically, they manifest as slight damage to the inter-turn insulation of the winding, resulting in contact resistance. The micro-short circuit path causes only a slight temperature change of 0.5~3℃ in the winding and a slight phase shift of 0°~5° in the three-phase current, without any obvious sudden change in current amplitude. It belongs to an early hidden fault that is extremely difficult to identify.

[0003] The operating status of high-voltage stator windings directly determines the safety, stability, and reliability of power supply in a power system. If minor inter-turn short circuits cannot be accurately identified in the early stages, the faults will continue to deteriorate and gradually evolve into serious inter-turn short circuits and phase-to-phase short circuits, eventually leading to unplanned shutdowns of power equipment, winding burnout, or even large-scale power grid outages, causing significant economic losses and safety risks. Therefore, early detection of such latent faults is crucial to ensuring the safe operation of power equipment.

[0004] Current winding fault detection methods are mainly divided into two categories: electrical feature detection and temperature feature detection. Both have significant technical shortcomings. Electrical feature-based methods rely on phase difference and sequence component analysis. However, they are easily overwhelmed by minor phase shifts caused by slight inter-turn short circuits, such as those caused by power frequency fluctuations, harmonic interference, and minor load changes. A single electrical feature cannot effectively distinguish between faults and interference, leading to significant missed detections. Temperature feature-based methods can only identify severe short-circuit faults with temperature surges ≥5℃. Furthermore, traditional contact-type temperature sensors are limited by the confined space inside high-voltage windings, making it difficult to accurately extract minute temperature difference signals (0.5~3℃). These methods also lack a quantitative correlation between temperature changes and electrical phase features, resulting in a disconnect between temperature and electrical characteristics and preventing dual-feature fusion for accurate judgment. In addition, existing detection methods often use a single threshold for judgment, lack a tiered early warning mechanism, are susceptible to instantaneous interference leading to misjudgments, and lack a real-time update mechanism for baseline features. This makes them unsuitable for adapting to long-term changes in winding operating conditions, resulting in insufficient engineering practicality and detection reliability.

[0005] To address the core challenges of limited temperature measurement in the confined space of high-voltage windings, difficulty in extracting minute temperature signals, susceptibility to interference from slight phase shifts, and the inability to deeply integrate temperature and electrical characteristics, and to overcome the technical bottleneck of traditional single-feature detection, a dual-feature identification method for minor inter-turn short circuits in high-voltage stator windings based on minute temperature signals is urgently needed to meet the engineering requirements for early, accurate, and stable identification of minor inter-turn short circuits in high-voltage stator windings. Summary of the Invention

[0006] To address the aforementioned problems in the prior art, this invention provides a dual-feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals. This method effectively solves the problems of insignificant temperature changes, difficulty in extracting minute temperature signals, and susceptibility to interference from slight phase shifts in minor inter-turn short circuit scenarios.

[0007] To achieve the above objectives, this invention proposes a dual-feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals, comprising: S1: The micro-temperature signal and three-phase current signal of the high-voltage stator winding are synchronously acquired using the same clock source to form a synchronous dataset. ,in for Time of the first Temperature values ​​of phase windings, for Current values ​​of phases A / B / C at time 1 / 2. =A / B / C; S2: The micro-temperature signal is sequentially processed by removing environmental and heat dissipation interference, calculating micro-temperature difference, and amplifying micro-signal to extract standardized temperature feature values. And locate the phase of minute temperature changes; S3: After preprocessing the three-phase current signals, extract the instantaneous phase of each phase current and calculate the micro-offset of the phase difference between any two phases. The phase difference standardized eigenvalues ​​are obtained after standardization. , And locate the phase difference micro-offset pairs that are associated with the micro-temperature changes; S4: Constructing a coupled feature vector based on the temperature influence mechanism of minor inter-turn short circuits in high-voltage stator windings, including standardized temperature eigenvalues, standardized phase difference eigenvalues, and temperature-phase difference correlation coefficients. ; S5: Based on the coupling feature vector Calculate the correlation degree of two-feature anomalies It identifies minor inter-turn short circuit fault types based on preset classification thresholds and executes corresponding fault response operations. S6: After completing the fault response, the temperature and phase difference characteristic data of this fault are included in the benchmark database, the benchmark temperature and benchmark phase difference under normal operating conditions are updated in real time, and the benchmark database is updated using a weighted average method.

[0008] Preferably, in step S1, the micro-temperature signal is acquired by using a non-contact infrared micro-temperature sensor array, arranged in the non-contact temperature measurement area on the outer side of the high-voltage stator winding slot, avoiding ventilation and heat dissipation interference points; the three-phase current acquisition uses phase accuracy... Rogowski coil at 0.1℃.

[0009] Preferably, in step S1, the parameters for synchronous acquisition using the same source clock are: clock synchronization is achieved using GPS or IEEE 1588 precise time protocol, the sampling frequency of the micro-temperature signal is 1Hz, the sampling frequency of the three-phase current signal is 20kHz, and the acquired data is uploaded to the detection terminal in real time for subsequent processing.

[0010] Preferably, in step S2, the specific method for eliminating environmental and heat dissipation interference is as follows: calculating the true temperature of the winding conductor by means of multiple sampling points plus background temperature difference method. The calculation formula is: ; in, for The ambient temperature of the high-voltage winding operating environment is constantly collected by an independent ambient temperature sensor. Air-cooled installation to improve the heat dissipation coefficient of the windings. =0.04, self-cooling installation =0.06.

[0011] Preferably, in step S2, the specific method for calculating the micro-temperature difference is as follows: calculate the micro-temperature difference between the actual temperature of each phase winding and the reference temperature under normal operating conditions. ; In the formula, The average temperature of the high-voltage stator winding during 72 hours of continuous fault-free operation; when At that time, the judgment of the first The phase is distinguished by slight temperature changes.

[0012] Preferably, in step S2, the specific method for amplifying the micro-signal is as follows: using a feature amplification method based on the temperature effect mechanism to amplify the micro-temperature difference. Convert to standardized temperature characteristic value The calculation formula is: ; in, This represents the maximum temperature variation value for a minor inter-turn short circuit.

[0013] Preferably, in S3, the phase difference micro-offset The calculation formula is: ; in, for time The instantaneous phase difference between the two phases, The average phase difference of the high-voltage stator winding during 72 hours of continuous fault-free operation; The formula for calculating the phase difference standardized eigenvalue is: ; in, =5° is the maximum phase difference micro-offset of a slight inter-turn short circuit.

[0014] Preferably, in step S4, the quantitative correlation formula for the temperature effect mechanism is: ; Coupled feature vectors The specific expression is: ; in, =0.00393 / ℃ is the temperature coefficient of resistance of the winding conductor. =0.0002 / ℃ is the temperature coefficient of the winding coil inductance. =314 rad / s is the power system angular frequency. This represents the DC resistance of the winding conductor at 20°C. The inductance of the winding coil at 20℃ These are the values ​​for minute temperature changes. This is a small phase difference offset. For the first With the Preprocessed phase difference characteristic signal between each winding For the first With the Preprocessed phase difference characteristic signal between each winding The temperature-phase difference correlation coefficient is calculated using the following formula: ; in, The covariance calculation function, The variance calculation function is used under the condition of slight inter-turn short circuit. ≥0.8.

[0015] Preferably, in step S5, the formula for calculating the correlation degree of dual-feature anomalies is... for: ; The grading determination threshold is: when and When, it is determined to be a normal operating condition; when and When, it is determined to be an early fault of minor inter-turn short circuit; when and At that time, it was determined to be a minor inter-turn short circuit confirmation fault.

[0016] Preferably, in S5, the corresponding fault response operation is as follows: under normal operating conditions, continuously monitor the operating status of the high-voltage stator winding; trigger a first-level early warning for a minor inter-turn short circuit early fault, continuously track changes in temperature and phase difference characteristics and record real-time data; trigger a second-level alarm for a minor inter-turn short circuit confirmed fault, upload fault information including fault phase, occurrence timestamp, and coupling characteristic data to the monitoring center, and start the high-voltage stator winding fault location process.

[0017] Therefore, this invention proposes a dual-feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals, with the following advantages: (1) The non-contact infrared micro-temperature sensor array is used to collect temperature, which is suitable for the installation limitation of the narrow space of the high voltage winding. With multi-dimensional interference elimination and micro-signal amplification processing, it can accurately extract weak temperature changes of 0.5~3℃, solving the core problems of traditional contact sensors being unable to be deployed and micro-temperature signals being easily submerged by noise.

[0018] (2) Based on the winding temperature influence mechanism, a quantitative correlation model of temperature and phase difference is established, and a dual feature coupling vector is constructed to achieve deep fusion, which completely breaks through the technical bottleneck of the disconnect between temperature features and electrical features, effectively avoids the problem of missed judgment in single feature detection, and can accurately identify early hidden inter-turn short circuits.

[0019] (3) The hierarchical judgment and real-time update mechanism of benchmark features are adopted to greatly reduce the misjudgment rate caused by instantaneous interference. No changes are required to the winding structure. It can be deployed with the help of general acquisition equipment. It is suitable for online monitoring scenarios of stator windings of high-voltage motors and generators. The fault identification accuracy is high and the engineering implementation is strong.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] Figure 1 This is an overall flowchart of the dual-feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals, as described in this invention. Figure 2 This is a flowchart of the micro-temperature signal processing method for the dual-feature identification method of slight inter-turn short circuit in high-voltage windings of the present invention; Figure 3 This is a flowchart of the coupled feature construction and fault classification judgment of the dual feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals of the present invention. Detailed Implementation

[0022] To make the technical solutions, advantages, and objectives of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below. The described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the protection scope of this application.

[0023] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0024] like Figures 1-3 As shown, the present invention provides a dual-feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals, comprising: S1: The micro-temperature signal and three-phase current signal of the high-voltage stator winding are synchronously acquired using the same clock source to form a synchronous dataset. ,in for Time of the first Temperature values ​​of phase windings, for Current values ​​of phases A / B / C at time 1 / 2. =A / B / C; The micro-temperature signal acquisition method is as follows: a non-contact infrared micro-temperature sensor array is used, arranged in the non-contact temperature measurement area on the outer side of the high-voltage stator winding slot, avoiding ventilation and heat dissipation interference points; the three-phase current acquisition adopts phase accuracy... Rogowski coil at 0.1℃.

[0025] The parameters for synchronous acquisition using the same source clock are as follows: clock synchronization is achieved using GPS or IEEE1588 precise time protocol, the sampling frequency of the micro-temperature signal is 1Hz, the sampling frequency of the three-phase current signal is 20kHz, and the acquired data is uploaded to the detection terminal in real time for subsequent processing.

[0026] S2: The micro-temperature signal is sequentially processed by removing environmental and heat dissipation interference, calculating micro-temperature differences, and amplifying the micro-signal to extract standardized temperature feature values. And locate the phase of minute temperature changes; The specific method for eliminating environmental and heat dissipation interference is as follows: calculate the true temperature of the winding conductor by means of multiple sampling points and background temperature difference method. The calculation formula is: ; in, for The ambient temperature of the high-voltage winding operating environment is constantly collected by an independent ambient temperature sensor. Air-cooled installation to improve the heat dissipation coefficient of the windings. =0.04, self-cooling installation =0.06.

[0027] The specific method for calculating the micro-temperature difference is as follows: calculate the micro-temperature difference between the actual temperature of each phase winding and the reference temperature under normal operating conditions. ; In the formula, The average temperature of the high-voltage stator winding during 72 hours of continuous fault-free operation; when At that time, the judgment of the first The phase is distinguished by slight temperature changes.

[0028] The specific method for amplifying micro signals is as follows: A feature amplification method based on the temperature effect mechanism is used to amplify micro-temperature differences. Convert to standardized temperature characteristic value The calculation formula is: ; in, This represents the maximum temperature variation value for a minor inter-turn short circuit.

[0029] S3: After preprocessing the three-phase current signals, extract the instantaneous phase of each phase current and calculate the micro-offset of the phase difference between any two phases. The phase difference standardized eigenvalues ​​are obtained after standardization. , And locate the phase difference micro-offset pairs that are associated with the micro-temperature changes; The preprocessing of the acquired three-phase current signals includes wavelet threshold denoising and hybrid morphological filtering. After preprocessing, an improved Hilbert-Huang transform is used to extract the instantaneous phase of each phase current. , , Calculate the phase difference between any two phases. , , Calculate the phase difference micro-offset , The formula is: Phase difference micro offset The calculation formula is: ; in, for time The instantaneous phase difference between the two phases, The average phase difference of the high-voltage stator winding during 72 hours of continuous fault-free operation; if The system determines that there is a slight phase shift and locates the phase shift pairs that are associated with slight temperature changes.

[0030] The process of obtaining the phase difference normalized eigenvalues ​​is as follows: based on the phase difference of minute temperature changes, extract its normalized temperature eigenvalues. Extract two sets of phase difference micro-offsets that are respectively correlated with micro-temperature changes. , The phase difference standardized eigenvalues ​​are then obtained through standardization. , The formula for calculating the standardized eigenvalues ​​of the phase difference is: ; in, For a two-phase combination associated with a slight temperature change, i.e. or , =5° is the maximum phase difference micro-offset of a slight inter-turn short circuit.

[0031] S4: Constructing a coupled feature vector based on the temperature influence mechanism of minor inter-turn short circuits in high-voltage stator windings, including standardized temperature eigenvalues, standardized phase difference eigenvalues, and temperature-phase difference correlation coefficients. ; The quantitative correlation formula for the temperature effect mechanism is as follows: ; Coupled feature vectors The specific expression is: ; in, =0.00393 / ℃ is the temperature coefficient of resistance of the winding conductor. =0.0002 / ℃ is the temperature coefficient of the winding coil inductance. =314 rad / s is the power system angular frequency. This represents the DC resistance of the winding conductor at 20°C. The inductance of the winding coil at 20℃ These are the values ​​for minute temperature changes. This is a small phase difference offset. For the first With the Preprocessed phase difference characteristic signal between each winding For the first With the Preprocessed phase difference characteristic signal between each winding The temperature-phase difference correlation coefficient is calculated using the following formula: ; in, The covariance calculation function, The variance calculation function is based on the temperature effect mechanism under the condition of slight inter-turn short circuit. ≥0.8 (strong positive correlation), under normal operating conditions <0.3 (no correlation).

[0032] S5: Based on the coupling feature vector Calculate the correlation degree of two-feature anomalies It identifies minor inter-turn short circuit fault types based on preset classification thresholds and executes corresponding fault response operations. Formula for calculating the correlation degree of dual-feature anomalies for: ; The weighting was verified by experiments, with temperature features accounting for 40% and phase difference features accounting for 60%. These were then multiplied by the correlation coefficient to ensure that the abnormal correlation was effective only when temperature and phase difference were strongly positively correlated.

[0033] The threshold for grading is: when and When this condition is deemed normal, continuous monitoring is maintained. when and When the fault is identified as an early stage of a minor inter-turn short circuit, a Level 1 warning is triggered, and the data is recorded. when and When the fault is confirmed as a minor inter-turn short circuit, a level-two alarm is triggered, and fault information (fault phase, timestamp, and characteristic data) is uploaded to the monitoring center.

[0034] The corresponding fault response operations are as follows: under normal operating conditions, continuously monitor the operating status of the high-voltage stator winding; a minor inter-turn short circuit early fault triggers a first-level early warning, continuously tracks changes in temperature and phase difference characteristics and records real-time data; a minor inter-turn short circuit confirmed fault triggers a second-level alarm, uploads fault information including fault phase, occurrence timestamp, and coupling characteristic data to the monitoring center, and initiates the high-voltage stator winding fault location process.

[0035] S6: After completing the fault response, the temperature and phase difference characteristic data of this fault are included in the benchmark database, the benchmark temperature and benchmark phase difference under normal operating conditions are updated in real time, and the benchmark database is updated using a weighted average method.

[0036] This invention takes the stator winding of a 35kV high-voltage asynchronous motor as an example to simulate a typical minor inter-turn short circuit condition (contact resistance ≥50Ω, corresponding to a slight temperature change of 0.5~3℃ and a slight phase shift of 0°~5°). This embodiment represents a mainstream equipment scenario in power systems and has high engineering applicability. The specific implementation process is as follows: I. Implementation Targets and Working Condition Simulation: The stator winding of a 35kV high-voltage asynchronous motor (rated voltage 35kV, rated power 10MW) was selected as the implementation carrier. By artificially simulating slight inter-turn insulation damage, a micro-short circuit with a contact resistance of 50Ω was set between slots 12 and 13 of phase A winding. The triggering conditions were as follows: the temperature of phase A winding increased by 1.8℃ compared with the normal operating conditions, the phase difference of current between phases A and B shifted by 3.2° compared with the reference value, the phase difference between phases BC and CA showed no significant shift, and there was no sudden change in current amplitude, which met the typical latent characteristics of a slight inter-turn short circuit.

[0037] II. Hardware Deployment and Signal Acquisition: Temperature sensor deployment: Four sets of non-contact infrared micro-temperature sensor arrays are arranged in the non-contact temperature measurement area outside the stator core slot and the end of the A-phase winding, avoiding heat dissipation interference points such as ventilation openings. Each array contains three sensing units to realize full-area temperature monitoring of the A-phase winding. The sampling frequency is set to 1Hz. Current acquisition setup: Install Rogowski coils with a phase accuracy of ±0.1° on the three-phase busbars A / B / C at the motor output terminals to acquire the three-phase current signals. The sampling frequency is set to 20kHz. Synchronization configuration: The GPS precise time protocol is used to achieve the same clock synchronization of temperature and current signals. The collected data is uploaded to the detection terminal in real time via industrial Ethernet to complete data preprocessing and storage.

[0038] III. Micro-temperature signal processing: Interference Removal: The detection terminal receives raw temperature data from four sensor arrays, removes instantaneous noise through multi-point mean filtering, and then calculates the true temperature of the winding conductor using the background temperature difference method. The calculation formula is as follows: ,in for The temperature of the high-voltage winding operating environment was measured at 25℃. =0.04 (heat dissipation coefficient under air-cooled conditions), the actual temperature of phase A winding is calculated to be 68.2℃; Micro-temperature difference calculation: The reference temperature obtained after 72 hours of trouble-free operation of the motor is retrieved. =66.4℃, calculate the micro temperature difference. Phase A was determined to be a phase with slight temperature changes. Micro-signal amplification: A temperature mechanism amplification method is used to convert micro-temperature differences into standardized temperature characteristic values. The calculation formula is as follows: ,in, Substituting into .

[0039] IV. Phase Difference Micro-Shift Feature Extraction: After wavelet denoising and morphological filtering to remove harmonic interference, the instantaneous phase of each phase was extracted by Hilbert-Huang transform. The phases of phase A, B, and C were measured to be 152.7°, 155.9°, and 149.3°, respectively. Calculate the phase difference between phases A and B. Adjust the reference phase difference The phase difference micro-offset is obtained. Standardized formula Calculation yields ; The correlation between phase difference shift and micro temperature change was verified, confirming that the phase difference shift between phases A and B corresponds to the micro temperature change of phase A.

[0040] V. Dual-feature fusion and fault diagnosis: A temperature-phase difference quantitative correlation model was constructed based on the temperature influence mechanism, and the temperature-phase difference correlation coefficient was calculated. =0.92 (≥0.8, meets the fault condition), construct the coupling feature vector; Calculate the correlation degree of two-feature anomalies Substituting into ; Based on the grading threshold: 2.0 ≤ 4.39 < 5.0 and =0.92≥0.8, which is determined to be an early stage of minor inter-turn short circuit fault in phase A, triggering a level one early warning, continuously tracking characteristic changes and recording real-time data.

[0041] VI. Baseline Updates and Subsequent Responses: After the fault is determined, the detection terminal will include the characteristic data of the slight temperature change and phase difference shift of phase A into the reference database, and update the reference temperature and reference phase difference under normal operating conditions using a weighted average method to adapt to the long-term operating conditions of the motor. At the same time, it will send an early warning information to the monitoring center, marking the faulty phase, characteristic parameters and early warning level. Maintenance personnel will carry out preventive maintenance based on the early warning information to avoid further deterioration of the fault.

[0042] This embodiment verifies the effectiveness of the method in the stator winding scenario of a 35kV high-voltage motor. It does not require modification of the original winding structure, can be deployed using general-purpose acquisition equipment, and can accurately identify early minor inter-turn short circuits. Its engineering practicality and detection reliability are outstanding.

[0043] Therefore, this invention provides a dual-feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals. This method solves the technical problems of limited space in high-voltage windings preventing the deployment of contact temperature sensors, difficulty in extracting minute temperature signals, susceptibility to interference from slight phase shifts, disconnect between temperature and electrical characteristics, and high false positive / false negative rates with single-feature detection. By combining non-contact temperature measurement with dual-feature coupling, it accurately identifies early-stage, latent inter-turn short circuits. The graded judgment and benchmark update mechanism significantly improves detection stability. This method requires no modification to the winding structure, can be deployed using general-purpose acquisition equipment, and is suitable for online monitoring of high-voltage motor and generator stator windings. It effectively avoids the risk of fault degradation, ensures the safe operation of power equipment, and significantly improves engineering practicality and reliability.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A dual-feature identification method for minor inter-turn short circuits in high-voltage windings based on minute temperature signals, wherein the minor inter-turn short circuit is caused by slight damage to the inter-turn / inter-strand insulation of the high-voltage stator winding, resulting in contact resistance. The micro-short-circuit path causes a micro-temperature change of 0.5~3°C in the winding and a micro-phase shift of 0°~5° in the three-phase current. Its characteristic is that... Includes the following steps: S1: The micro-temperature signal and three-phase current signal of the high-voltage stator winding are synchronously acquired using the same clock source to form a synchronous dataset. ,in for Time of the first Temperature values ​​of phase windings, for Current values ​​of phases A / B / C at time 1 / 2. =A / B / C; S2: The micro-temperature signal is sequentially processed by removing environmental and heat dissipation interference, calculating micro-temperature difference, and amplifying micro-signal to extract standardized temperature feature values. And locate the phase of minute temperature changes; S3: After preprocessing the three-phase current signals, extract the instantaneous phase of each phase current and calculate the micro-offset of the phase difference between any two phases. The phase difference standardized eigenvalues ​​are obtained after standardization. , And locate the phase difference micro-offset pairs that are associated with the micro-temperature changes; S4: Constructing a coupled feature vector based on the temperature influence mechanism of minor inter-turn short circuits in high-voltage stator windings, including standardized temperature feature values, standardized phase difference feature values, and temperature-phase difference correlation coefficients. ; S5: Based on the coupling feature vector Calculate the correlation degree of two-feature anomalies It identifies minor inter-turn short circuit fault types based on preset classification thresholds and executes corresponding fault response operations. S6: After completing the fault response, the temperature and phase difference characteristic data of this fault are included in the benchmark database, the benchmark temperature and benchmark phase difference under normal operating conditions are updated in real time, and the benchmark database is updated using a weighted average method.

2. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 1, characterized in that, In S1, the micro-temperature signal is acquired by using a non-contact infrared micro-temperature sensor array, arranged in the non-contact temperature measurement area on the outside of the high-voltage stator winding slot and end, avoiding ventilation and heat dissipation interference points; the three-phase current acquisition uses phase accuracy... Rogowski coil at 0.1℃.

3. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 2, characterized in that, In S1, the parameters for synchronous acquisition of the same source clock are as follows: clock synchronization is achieved using GPS or IEEE1588 precise time protocol, the sampling frequency of micro-temperature signal is 1Hz, the sampling frequency of three-phase current signal is 20kHz, and the acquired data is uploaded to the detection terminal in real time for subsequent processing.

4. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 3, characterized in that, In step S2, the specific method for eliminating environmental and heat dissipation interference is as follows: the true temperature of the winding conductor is calculated by multi-sampling point mean filtering plus background temperature difference method. The calculation formula is: ； in, for The ambient temperature of the high-voltage winding operating environment is constantly collected by an independent ambient temperature sensor. Air-cooled installation to improve the heat dissipation coefficient of the windings. =0.04, self-cooling installation =0.

06.

5. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 4, characterized in that, In step S2, the specific method for calculating the micro-temperature difference is as follows: calculate the micro-temperature difference between the actual temperature of each phase winding and the reference temperature under normal operating conditions. ； In the formula, The average temperature of the high-voltage stator winding during 72 hours of continuous fault-free operation; when At that time, the judgment of the first The phase is distinguished by slight temperature changes.

6. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 5, characterized in that, In S2, the specific method for amplifying the micro-signal is as follows: A feature amplification method based on the temperature effect mechanism is used to amplify the micro-temperature difference. Convert to standardized temperature characteristic value The calculation formula is: ； in, This represents the maximum temperature variation value for a minor inter-turn short circuit.

7. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 6, characterized in that, In S3, the phase difference micro-offset The calculation formula is: ； in, for time The instantaneous phase difference between the two phases, The average phase difference of the high-voltage stator winding during 72 hours of continuous fault-free operation; The formula for calculating the phase difference standardized eigenvalue is: ； in, =5° is the maximum phase difference micro-offset of a slight inter-turn short circuit.

8. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 7, characterized in that, In S4, the quantitative correlation formula for the temperature effect mechanism is as follows: ； Coupled feature vectors The specific expression is: ； in, =0.00393 / ℃ is the temperature coefficient of resistance of the winding conductor. =0.0002 / ℃ is the temperature coefficient of the winding coil inductance. =314 rad / s is the power system angular frequency. This represents the DC resistance of the winding conductor at 20°C. The inductance of the winding coil at 20℃ These are the values ​​for minute temperature changes. This is a small phase difference offset. For the first With the Preprocessed phase difference characteristic signal between each winding For the first With the Preprocessed phase difference characteristic signal between each winding The temperature-phase difference correlation coefficient is calculated using the following formula: ； in, The covariance calculation function, The variance calculation function is used under the condition of slight inter-turn short circuit. ≥0.

8.

9. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 8, characterized in that, In S5, the formula for calculating the correlation degree of dual-feature anomalies is... for: ； The grading determination threshold is: when and When, it is determined to be a normal operating condition; when and When, it is determined to be an early fault of minor inter-turn short circuit; when and At that time, it was determined to be a minor inter-turn short circuit confirmation fault.

10. The dual-feature identification method for slight inter-turn short circuits in high-voltage windings based on minute temperature signals according to claim 9, characterized in that, In S5, the corresponding fault response operation is as follows: under normal operating conditions, continuously monitor the operating status of the high-voltage stator winding; trigger a first-level early warning for a minor inter-turn short circuit early fault, continuously track changes in temperature and phase difference characteristics and record real-time data; trigger a second-level alarm for a minor inter-turn short circuit confirmed fault, upload fault information including fault phase, occurrence timestamp, and coupling characteristic data to the monitoring center, and start the high-voltage stator winding fault location process.

Images