Dielectric loss detection method for high and low voltage mutual inductor
Through multi-stage voltage application strategies, spectrum analysis and artificial intelligence self-learning algorithms and other technologies, combined with environmental preprocessing and data visualization, the problem that traditional media-destruction and data visualization is solved, and high-precision fault prediction and maintenance suggestions are achieved, improving the Testing efficiency and equipment management efficiency.
Patent Information
- Application Number
- CN202411512707.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-28
- Publication Date
- 2024-11-26
AI Technical Summary
Traditional dilemma detection methods cannot dynamically adjust the test plan according to the actual operating status of the transformer and changes in the external environment, resulting in insufficient accuracy and inefficiency of the detection results, lack of effective trend analysis and prediction capabilities, resulting in lag and passive maintenance work.
Multi-stage voltage application strategy, spectrum analysis, artificial intelligence self-learning algorithm and multi-source data fusion technology are adopted, combined with environmental preprocessing and data visualization, detection parameters are dynamically adjusted, long-term dielectric loss model is constructed, operating trends are predicted, and maintenance suggestions are generated.
Achieve high-precision hidden fault identification and future risk prediction, improve detection accuracy and efficiency, promote preventive maintenance, extend equipment service life, and improve maintenance efficiency and safety through automated reporting and remote monitoring. .
Smart Images

Figure CN119024055A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of transformer dielectric loss detection, and in particular to a dielectric loss detection method for high and low voltage transformers. Background Art
[0002] The dielectric loss of a mutual inductance transformer is generally characterized by insulation sensitive parameters. Depolarization current prediction can be used to estimate key insulation sensitive parameters, such as depolarization charge distribution. By analyzing the depolarization current data, key information about the year-sensitive parameters of the mutual inductance transformer OPI operating status can be extracted.
[0003] Traditional dielectric loss detection methods usually use fixed test parameters for measurement and are unable to dynamically adjust the test scheme according to the actual operating status of the transformer and changes in the external environment. This one-size-fits-all detection method is prone to problems such as insufficient detection accuracy or low test efficiency when facing transformers of different models and working conditions.
[0004] Existing dielectric loss detection methods mainly focus on detecting the current status of equipment and lack effective trend analysis and prediction capabilities. Due to the lack of advance prediction of future operating risks, maintenance personnel can often only perform maintenance after faults appear or equipment performance has significantly declined, resulting in delayed and passive maintenance work.
[0005] In view of these deficiencies in the prior art, the present invention proposes a dielectric loss detection method for high and low voltage transformers. Summary of the invention
[0006] In view of the shortcomings of the prior art, the present invention provides a dielectric loss detection method for high and low voltage transformers, which solves the problem that traditional dielectric loss detection methods usually use fixed test parameters for measurement and cannot dynamically adjust the test scheme according to the actual operating status of the transformer and changes in the external environment.
[0007] To achieve the above objectives, the present invention is implemented by the following technical solutions: A dielectric loss detection method for high and low voltage transformers, comprising the following steps: Perform environmental pretreatment on the transformer, adjust the surface and internal conditions based on temperature, humidity and air pressure factors, and preheat to ensure stability; Adopt multi-stage voltage application strategy, dynamically monitor transient response, obtain leakage current and dielectric loss data, and perform frequency scanning; Perform spectrum analysis on dielectric loss data, combine frequency domain and time domain methods to extract key eigenvalues; Apply artificial intelligence self-learning algorithms to optimize test parameters in real time and dynamically adjust detection plans; Utilize multi-source data fusion technology to integrate test data and historical data, build a long-term dielectric loss model, and predict operating trends; The analysis results are compared with the standard model to generate a dielectric loss detection report, identify hidden faults and provide maintenance recommendations.
[0008] Preferably, the preprocessing includes using high-precision sensors to monitor environmental conditions in real time, specifically including temperature, humidity, air pressure and electromagnetic interference factors, and dynamically adjusting the heating and voltage stabilization process of the transformer according to the detected environmental data, and adjusting the preheating time in real time in combination with environmental changes to ensure that the transformer is in the most stable and suitable test state.
[0009] Preferably, the voltage application strategy adopts a progressive voltage waveform, combined with a gradual increase in voltage in each stage, monitors the insulation layer reaction of the transformer in real time during the voltage change process, and implements a transition voltage in each voltage stage to ensure the continuity and reliability of the dielectric loss measurement, while reducing potential damage to the insulation structure of the transformer during the test.
[0010] Preferably, the spectrum analysis is combined with wavelet transform and Fourier transform to decompose the dielectric loss data to ensure accurate capture of high-frequency and low-frequency characteristics.
[0011] Preferably, the characteristic values include high-frequency and low-frequency limit values of the dielectric loss factor and their change rates, which are used to evaluate the insulation condition, wherein the change rate can be expressed as:
[0012] in, Indicates at time The dielectric loss factor measured at all times, Indicates that at the next time point The dielectric loss factor measured at all times, is the rate of change of dielectric loss factor.
[0013] Preferably, the self-learning algorithm combines operational data and historical data to automatically adjust test voltage, frequency and time to improve detection efficiency and reduce electrical stress.
[0014] Preferably, the data fusion technology combines current detection data, historical data and comparison data of similar transformers to improve the accuracy and reliability of the prediction model. The data fusion process also includes outlier removal and data smoothing to eliminate possible measurement noise and data deviation, thereby providing a more stable prediction result.
[0015] Preferably, the constructed long-period dielectric loss model adopts a machine learning algorithm and is optimized through continuously updated data to adapt to characteristic changes in different operating stages.
[0016] Preferably, the test report intuitively displays the health status and future operating risks of the transformer through data visualization technology. The test report also integrates an automated fault analysis module, which can automatically generate maintenance suggestions by comparing with historical data, and issue alarms and maintenance notifications to relevant maintenance personnel through a remote monitoring system.
[0017] Preferably, the maintenance suggestion is based on the future trend calculated by the following formula:
[0018] in, For time The predicted dielectric loss factor value at time is the baseline dielectric loss factor value, is the initial offset, is the dielectric loss factor attenuation rate constant, For time.
[0019] The present invention provides a dielectric loss detection method for high and low voltage transformers. It has the following beneficial effects: 1. By combining spectrum analysis, historical data comparison and future trend prediction, the present invention can not only accurately identify potential hidden faults of transformers in current operation, but also predict future risks in advance based on trend analysis. The maintenance suggestions automatically generated by the fault analysis module can effectively guide preventive maintenance, avoid unplanned downtime caused by faults, and further extend the service life of the equipment. This intelligent fault prediction and maintenance suggestion mechanism makes maintenance work more forward-looking and efficient.
[0020] 2. The present invention realizes dynamic optimization of parameters and adaptive adjustment of test schemes during the detection process by adopting an artificial intelligence-based self-learning algorithm and multi-source data fusion technology. The method can automatically adjust test parameters such as voltage and frequency according to the real-time operating status and environmental conditions of the transformer, thereby improving the accuracy and efficiency of detection. At the same time, the introduction of data fusion technology enables the detection model to continuously learn and optimize, and can better adapt to the changes in characteristics of different operating stages, ensuring that stable and reliable detection results can be obtained under complex working conditions.
[0021] 3. The present invention uses data visualization technology to display the test results in an intuitive graphical manner, making the equipment status and operating risks clear at a glance, which is convenient for operators to quickly understand and make decisions. In addition, the automatic generation of test reports and the integration of remote monitoring systems ensure that relevant maintenance personnel can obtain key information and take actions in a timely manner, thereby significantly improving the efficiency of equipment management. Through the real-time monitoring and automatic alarm mechanism of the system, the complexity and response time of maintenance work are reduced, providing a strong guarantee for the safe operation of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION
[0023] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0024] Example: Please see attached Figure 1 The embodiment of the present invention provides a dielectric loss detection method for high and low voltage transformers, comprising the following steps: Perform environmental pretreatment on the transformer, adjust the surface and internal conditions based on temperature, humidity and air pressure factors, and preheat to ensure stability; Adopt multi-stage voltage application strategy, dynamically monitor transient response, obtain leakage current and dielectric loss data, and perform frequency scanning; Perform spectrum analysis on dielectric loss data, combine frequency domain and time domain methods, and extract key eigenvalues; Apply artificial intelligence self-learning algorithms to optimize test parameters in real time and dynamically adjust detection plans; Utilize multi-source data fusion technology to integrate test data and historical data, build a long-term dielectric loss model, and predict operating trends; The analysis results are compared with the standard model to generate a dielectric loss detection report, identify hidden faults and provide maintenance recommendations.
[0025] Specifically, the transformer is subjected to environmental pretreatment, and the surface and internal conditions of the transformer are adjusted by comprehensively considering factors such as temperature, humidity and air pressure. The transformer is also preheated to ensure that it is in a stable state during the test. Subsequently, a multi-stage voltage application strategy is adopted to dynamically monitor the transient response of the transformer at each voltage stage, obtain real-time leakage current and dielectric loss data, and perform frequency scanning at the same time to fully capture the dielectric loss characteristics of the transformer under different voltage and frequency conditions. The acquired dielectric loss data is subjected to spectral analysis, and the key eigenvalues are deeply extracted by combining frequency domain and time domain methods to identify the dielectric loss characteristics that may affect the performance of the transformer. Artificial intelligence is applied to identify the dielectric loss characteristics that may affect the performance of the transformer. Intelligent self-learning algorithm optimizes test parameters and detection processes in real time, and dynamically adjusts detection schemes to adapt to the performance of the transformer under different operating states and environmental conditions, ensuring the accuracy and efficiency of detection; at the same time, multi-source data fusion technology is used to integrate current detection data with historical data to build a long-term dielectric loss model for the transformer. The model is used to predict the future operating trend of the transformer and provide data support; finally, the analysis results are compared with the standard model to generate a multi-dimensional dielectric loss detection report, identify potential hidden faults or abnormal trends, and put forward targeted maintenance suggestions to ensure the long-term safe and reliable operation of the transformer.
[0026] The preprocessing includes using high-precision sensors to monitor environmental conditions in real time, specifically including temperature, humidity, air pressure and electromagnetic interference factors, and dynamically adjusting the heating and voltage stabilization process of the transformer according to the detected environmental data, and adjusting the preheating time in real time in combination with environmental changes to ensure that the transformer is in the most stable and suitable test state.
[0027] Specifically, the factors monitored include external environmental factors such as temperature, humidity, air pressure, and electromagnetic interference. Based on these detected environmental data, the system dynamically adjusts the heating and voltage stabilization process of the transformer to ensure that the transformer is tested under appropriate conditions. In addition, combined with real-time monitoring of environmental changes, the system will automatically adjust the preheating time to ensure that the transformer is always in the most stable and suitable test state, thereby minimizing the impact of the external environment on the test results and improving the accuracy and reliability of the detection.
[0028] The voltage application strategy adopts a progressive voltage waveform, combined with a gradual increase in voltage at each stage, monitors the insulation layer reaction of the transformer in real time during the voltage change process, and implements a transition voltage at each voltage stage to ensure the continuity and reliability of dielectric loss measurement, while reducing potential damage to the insulation structure of the transformer during the test.
[0029] Specifically, the voltage application strategy uses a progressive voltage waveform to ensure that the transformer is tested step by step by gradually increasing the voltage. At each stage of voltage change, the response of the transformer insulation layer is monitored in real time to capture the dynamic changes in insulation performance. In order to ensure the continuity and reliability of dielectric loss measurement, a transition voltage step is introduced at each voltage stage to make the voltage smoothly transition from one level to the next, avoiding stress concentration on the insulation layer caused by sudden voltage changes, thereby reducing potential damage to the transformer insulation structure during the test. This strategy can not only ensure the stability of the measurement data, but also effectively protect the long-term operation safety of the transformer.
[0030] The spectrum analysis combines wavelet transform and Fourier transform to decompose the dielectric loss data to ensure accurate capture of high-frequency and low-frequency characteristics.
[0031] Specifically, first, the dielectric loss data is converted to the frequency domain using Fourier transform to analyze its global frequency characteristics; then, the data is subjected to fine time-frequency analysis through wavelet transform, so as to accurately capture the high-frequency and low-frequency characteristics of the data and the local characteristics of the signal at different time scales. This method ensures that when performing a comprehensive spectrum analysis of the dielectric loss data, not only the overall frequency distribution can be extracted, but also the key frequency components in the dielectric loss characteristics of the transformer can be deeply identified, thereby providing more accurate and detailed data support for subsequent fault diagnosis and trend prediction.
[0032] The characteristic values include the high-frequency and low-frequency limit values of the dielectric loss factor and their change rates, which are used to evaluate the insulation condition, wherein the change rate can be expressed as:
[0033] in, Indicates at time The dielectric loss factor measured at all times, Indicates that at the next time point The dielectric loss factor measured at all times, is the rate of change of dielectric loss factor.
[0034] Specifically, by calculating these change rates, the changing trend of the transformer insulation performance over time can be evaluated, helping to identify potential signs of degradation and hidden faults, thereby providing a scientific basis for timely preventive maintenance measures.
[0035] The self-learning algorithm combines operational and historical data to automatically adjust test voltage, frequency and time to improve detection efficiency and reduce electrical stress.
[0036] Specifically, the current operating status of the transformer is analyzed first, and the patterns and trends in the historical data are used to automatically optimize the application value of the test voltage and the selection of the frequency to ensure the accuracy of the test under different working conditions. In addition, the algorithm automatically adjusts the test time and test sequence based on the electrical stress feedback monitored in real time to reduce the stress impact on the electrical insulation of the transformer. In this way, the self-learning algorithm can improve detection efficiency, reduce unnecessary repeated tests, and effectively extend the service life of the transformer, ensuring that high-quality measurement data is obtained under optimal conditions.
[0037] The data fusion technology combines current detection data, historical data and comparison data of similar transformers to improve the accuracy and reliability of the prediction model. The data fusion process also includes outlier removal and data smoothing to eliminate possible measurement noise and data deviation, thereby providing more stable prediction results.
[0038] Specifically, data from different sources are integrated, and through multi-dimensional analysis, the current detection results are deeply compared with historical trends and data from similar devices to enhance the adaptability of the prediction model to various operating conditions. At the same time, outlier removal and data smoothing steps are introduced in the data fusion process to eliminate possible measurement noise by identifying and removing outliers, and smooth the data to reduce prediction deviations caused by data fluctuations. After such processing, the generated prediction results are more stable and reliable, providing a more credible basis for the status assessment and future operating trends of the transformer, thereby effectively guiding maintenance and management decisions.
[0039] The constructed long-cycle dielectric loss model adopts a machine learning algorithm and is optimized through continuously updated data to adapt to characteristic changes in different operating stages.
[0040] Specifically, the model is continuously optimized using constantly updated test data. The model automatically identifies the changing patterns of the dielectric loss characteristics during operation by learning the historical data and real-time monitoring data of the transformer at different operating stages. As the test data increases, the machine learning algorithm can dynamically adjust the model parameters to adapt to the changes in the characteristics of the transformer under different load, environmental and time conditions. Through this adaptive optimization process, the model can not only accurately predict the future dielectric loss trend of the transformer, but also respond to any changes in the transformer status in a timely manner, providing more accurate status assessment and prediction results, thereby providing a scientific basis for formulating maintenance strategies and ensuring the long-term stable operation of the transformer.
[0041] The test report intuitively displays the health status and future operating risks of the transformer through data visualization technology. The test report also integrates an automated fault analysis module, which can automatically generate maintenance suggestions by comparing with historical data, and send alarms and maintenance notifications to relevant maintenance personnel through a remote monitoring system.
[0042] Specifically, the health status and future operating risks of the transformer are displayed in intuitive forms such as charts, trend lines and heat maps, so that relevant personnel can quickly understand the current status and potential problems of the equipment. The report not only includes the display of static data, but also integrates an automated fault analysis module, which can analyze the differences between current detection data and historical data in real time, and automatically generate detailed maintenance recommendations by identifying abnormal trends or indicators that deviate from the normal range. Furthermore, the detection report is also connected to the maintenance team through a remote monitoring system. When the system detects potential failure risks or situations that require emergency treatment, it automatically triggers an alarm and sends a maintenance notification to ensure that relevant maintenance personnel can receive information in a timely manner and take necessary preventive or repair measures, thereby greatly improving the operating safety and management efficiency of the transformer.
[0043] The maintenance recommendations are based on future trends calculated using the following formula:
[0044] in, For time The predicted dielectric loss factor value at time is the baseline dielectric loss factor value, is the initial offset, is the dielectric loss factor attenuation rate constant, For time.
[0045] Specifically, by fitting current and historical data, the change trend of the transformer's dielectric loss in the future is predicted. Based on this trend, the system automatically generates targeted maintenance suggestions to help maintenance personnel predict the performance changes of the transformer and take measures in advance to prevent potential failures and ensure the long-term stable operation of the equipment.
[0046] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A dielectric loss detection method for high and low voltage transformers, characterized in that: The following steps are involved: Perform environmental pretreatment on the transformer, adjust the surface and internal conditions based on temperature, humidity and air pressure factors, and preheat to ensure stability; Adopt multi-stage voltage application strategy, dynamically monitor transient response, obtain leakage current and dielectric loss data, and perform frequency scanning; Perform spectrum analysis on dielectric loss data, combine frequency domain and time domain methods to extract key eigenvalues; Apply artificial intelligence self-learning algorithms to optimize test parameters in real time and dynamically adjust detection plans; Utilize multi-source data fusion technology to integrate test data and historical data, build a long-term dielectric loss model, and predict operating trends; The analysis results are compared with the standard model to generate a dielectric loss detection report, identify hidden faults and provide maintenance recommendations.
2. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The preprocessing includes using high-precision sensors to monitor environmental conditions in real time, specifically including temperature, humidity, air pressure and electromagnetic interference factors, and dynamically adjusting the heating and voltage stabilization process of the transformer according to the detected environmental data, and adjusting the preheating time in real time in combination with environmental changes to ensure that the transformer is in the most stable and suitable test state.
3. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The voltage application strategy adopts a progressive voltage waveform, combined with a gradual increase in voltage at each stage, monitors the insulation layer reaction of the transformer in real time during the voltage change process, and implements a transition voltage at each voltage stage to ensure the continuity and reliability of dielectric loss measurement, while reducing potential damage to the insulation structure of the transformer during the test.
4. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The spectrum analysis combines wavelet transform and Fourier transform to decompose the dielectric loss data to ensure accurate capture of high-frequency and low-frequency characteristics.
5. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The characteristic values include the high-frequency and low-frequency limit values of the dielectric loss factor and their change rates, which are used to evaluate the insulation condition, wherein the change rate can be expressed as: in, Indicates at time The dielectric loss factor measured at all times, Indicates that at the next time point The dielectric loss factor measured at all times, is the rate of change of dielectric loss factor.
6. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The self-learning algorithm combines operational and historical data to automatically adjust test voltage, frequency and time to improve detection efficiency and reduce electrical stress.
7. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The data fusion technology combines current detection data, historical data and comparison data of similar transformers to improve the accuracy and reliability of the prediction model. The data fusion process also includes outlier removal and data smoothing to eliminate possible measurement noise and data deviation, thereby providing more stable prediction results.
8. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The constructed long-cycle dielectric loss model adopts a machine learning algorithm and is optimized through continuously updated data to adapt to characteristic changes in different operating stages.
9. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The test report intuitively displays the health status and future operating risks of the transformer through data visualization technology. The test report also integrates an automated fault analysis module, which can automatically generate maintenance suggestions by comparing with historical data, and send alarms and maintenance notifications to relevant maintenance personnel through a remote monitoring system.
10. A dielectric loss detection method for high and low voltage transformers according to claim 1, characterized in that: The maintenance recommendations are based on future trends calculated using the following formula: in, For time The predicted dielectric loss factor value at time is the baseline dielectric loss factor value, is the initial offset, is the dielectric loss factor attenuation rate constant, For time.
