Transformer oil chromatography monitoring method and system, intelligent terminal and storage medium

Abstract

The invention relates to a transformer oil chromatography monitoring method and system, an intelligent terminal and a storage medium, and relates to the field of intelligent monitoring. Retrieving first transformer temperature prediction time sequence information meeting the transformer control parameter time sequence matrix and the environment monitoring temperature; if yes, processing the transformer control parameter time sequence matrix through a temperature predictor bound with the transformer bit number to obtain second transformer temperature prediction time sequence information; according to first or second transformer temperature prediction time sequence information, centroid transformer oil chromatograms of the same kind of transformer oil healthy sample group are retrieved in combination with transformer oil service duration, and a reference substance type list and a reference content interval list are obtained to execute transformer oil chromatogram monitoring; the technical problem that monitoring is inaccurate due to the fact that transformer oil chromatography monitoring is difficult to adapt to a dynamically changing thermoelectric environment is solved.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring, and in particular to a method, system, intelligent terminal, and storage medium for monitoring transformer oil chromatography. Background Technology

[0002] In the operation of power systems, transformers are critical equipment, and the stability and reliability of their operation are essential for the safety of the entire system. Transformer oil, as a vital insulating and cooling medium, directly affects the transformer's operating efficiency and lifespan. Therefore, regular monitoring and analysis of transformer oil, to promptly identify and resolve potential problems, is a crucial means of ensuring the normal operation of transformers.

[0003] Existing methods for monitoring transformer oil using chromatography primarily rely on periodic sampling of the transformer oil, followed by laboratory analysis to quantitatively and qualitatively determine the types and contents of substances. While this method can reflect the state of the transformer oil to some extent, it has several shortcomings. First, since the formation of substances is mainly due to the effects of heat and electricity, and the thermoelectric environment is easily affected by various external factors in practical applications, such as climate change and equipment aging, the changes in the substances in the transformer oil are not entirely determined by the transformer's own condition, making it difficult to accurately reflect the true state of the transformer oil. Second, existing technologies rely on fixed types and contents for decision-making, lacking dynamic adaptability to changes in the state of transformer oil under different environmental and operating conditions. In practical applications, the state of transformer oil may vary significantly under different transformers, different operating environments, and different time periods. Therefore, fixed types and contents are insufficient to meet monitoring needs under different circumstances. Furthermore, this method suffers from long monitoring cycles and slow response times. The need for periodic sampling and laboratory analysis makes it difficult to detect abnormal changes in transformer oil in a timely manner, potentially leading to delays in addressing potential faults.

[0004] In summary, existing transformer oil chromatographic monitoring methods have shortcomings in terms of accuracy, dynamic adaptability, and real-time monitoring capabilities, making it difficult to meet the high requirements of modern power systems for transformer oil condition monitoring. Therefore, a new type of transformer oil chromatographic monitoring method is needed that can more accurately and flexibly reflect the condition of transformer oil and provide timely early warning of potential faults. Summary of the Invention

[0005] This invention addresses the technical problem in existing transformer oil chromatography monitoring technologies that struggle to adapt to dynamically changing thermoelectric environments, leading to inaccurate monitoring. It provides a transformer oil chromatography monitoring method, system, intelligent terminal, and storage medium to solve this problem.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: In a first aspect, the present invention provides a method for monitoring transformer oil chromatography, comprising: when the environmental temperature deviation vector between the environmental monitoring temperature and the reference environmental temperature is greater than or equal to an environmental temperature deviation threshold, retrieving first transformer temperature prediction timing information that satisfies the transformer control parameter timing matrix and the environmental monitoring temperature; when the environmental temperature deviation vector between the environmental monitoring temperature and the reference environmental temperature is less than the environmental temperature deviation threshold, processing the transformer control parameter timing matrix through a temperature predictor bound to the transformer tag number to obtain second transformer temperature prediction timing information; based on the first transformer temperature prediction timing information or the second transformer temperature prediction timing information, combined with the transformer oil service life, retrieving centroid transformer oil chromatography data of similar healthy transformer oil sample groups to obtain a reference substance type list and a reference content range list, and performing transformer oil chromatography monitoring.

[0007] Secondly, the present invention provides a transformer oil chromatography monitoring system, the system comprising: an information retrieval module, used to retrieve first transformer temperature prediction timing information that satisfies the transformer control parameter timing matrix and the environmental monitoring temperature when the environmental temperature deviation vector between the environmental monitoring temperature and the reference environmental temperature is greater than or equal to an environmental temperature deviation threshold; a data processing module, used to process the transformer control parameter timing matrix through a temperature predictor bound to the transformer tag number to obtain second transformer temperature prediction timing information when the environmental temperature deviation vector between the environmental monitoring temperature and the reference environmental temperature is less than the environmental temperature deviation threshold; and a monitoring execution module, used to retrieve centroid transformer oil chromatography data of similar healthy transformer oil sample groups based on the first transformer temperature prediction timing information or the second transformer temperature prediction timing information, combined with the transformer oil service life, to obtain a list of reference substance types and a list of reference content ranges, and to perform transformer oil chromatography monitoring.

[0008] Thirdly, the present invention provides a smart terminal, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby realizing a transformer oil chromatography monitoring method as described in the first aspect.

[0009] Fourthly, the present invention provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements a transformer oil chromatographic monitoring method as described in the first aspect.

[0010] The beneficial effects of this invention are: by selecting different temperature prediction strategies based on the environmental temperature deviation, and by combining the service life of transformer oil to search for centroid transformer oil chromatograms of similar healthy sample groups, transformer oil chromatographic monitoring can be performed more accurately, thereby improving the accuracy and adaptability of monitoring. Attached Figure Description

[0011] Figure 1 This is a schematic flowchart of a transformer oil chromatographic monitoring method provided by the present invention.

[0012] Figure 2 This is a schematic diagram of the structure of a transformer oil chromatography monitoring system provided by the present invention.

[0013] Figure 3 This is a schematic diagram of the structure of the smart terminal provided by the present invention.

[0014] Figure 4 This is a schematic diagram of the structure of a computer-readable storage medium provided by the present invention.

[0015] Explanation of reference numerals in the attached drawings: Information retrieval module 11, data processing module 12, monitoring and execution module 13, intelligent terminal 500, memory 510, processor 520, first computer program 511, computer-readable storage medium 600, second computer program 611. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0018] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein. Example

[0019] like Figure 1 As shown, this embodiment of the invention provides a method for chromatographic monitoring of transformer oil, including: S10: When the environmental temperature deviation vector between the environmental monitoring temperature and the reference environmental temperature is greater than or equal to the environmental temperature deviation threshold, retrieve the first transformer temperature prediction timing information that satisfies the transformer control parameter timing matrix and the environmental monitoring temperature.

[0020] S20: When the ambient temperature deviation vector between the monitored ambient temperature and the reference ambient temperature is less than the ambient temperature deviation threshold, the temperature predictor bound to the transformer tag number processes the timing matrix of the transformer control parameters to obtain the timing information of the second transformer temperature prediction.

[0021] S30: Based on the first transformer temperature prediction time series information or the second transformer temperature prediction time series information, and combined with the transformer oil service life, retrieve the centroid transformer oil chromatography data of similar healthy transformer oil sample groups, obtain a list of reference substance types and a list of reference content ranges, and perform transformer oil chromatography monitoring.

[0022] For example, this solution determines the appropriate temperature prediction strategy based on the deviation between the monitored ambient temperature and the reference ambient temperature during transformer oil chromatography monitoring. Ambient temperature refers to the currently monitored external ambient temperature of the transformer, which is a constantly changing parameter influenced by factors such as season, weather, and geographical location. The reference ambient temperature, on the other hand, is a fixed temperature value used for comparison and reference. Determined based on historical data, industry standards, or the operating experience of a specific transformer, the reference ambient temperature represents a normal or expected ambient temperature level. It is used as the basis for judging whether the current ambient temperature is abnormal and, accordingly, determining the appropriate temperature prediction strategy.

[0023] When the deviation between the monitored ambient temperature and the reference ambient temperature reaches or exceeds the set ambient temperature deviation threshold, a specific temperature prediction process is initiated. This involves first retrieving transformer temperature monitoring time-series information that matches the current monitored ambient temperature and meets the requirements of the transformer control parameter time-series matrix. The transformer control parameter time-series matrix contains various operating parameters of the transformer and their changes over time. The transformer temperature monitoring time-series information is typically derived from historical monitoring data, which is selected to ensure it accurately reflects the temperature changes of the transformer under different ambient temperatures and control parameters. Subsequently, cluster analysis is performed on the retrieved information to identify patterns and trends in the data. During this process, some abnormal or unexpected data points may be removed to ensure the accuracy and reliability of the prediction results. Finally, based on the results of the cluster analysis, the temporal temperature distribution range of the retained data is statistically determined. This range is the first transformer temperature prediction time-series information, representing the possible range of transformer temperature changes under the current ambient temperature and control parameters. The retained data refers to data identified during the cluster analysis process as meeting specific conditions or patterns and therefore retained for further analysis. Once a set of transformer temperature monitoring time-series information that satisfies the transformer control parameter time-series matrix and environmental monitoring temperature is retrieved, cluster analysis is performed. During this process, some data points or clusters may be found to differ significantly from the overall dataset. These may represent abnormal operating conditions or erroneous monitoring results. These data points or clusters are typically considered noise or outliers and are removed during cluster analysis. The retained data, identified as consistent with the overall dataset and accurately reflecting the transformer's temperature variation under different environmental temperatures and control parameters, is used to statistically analyze the temperature distribution range at any given moment, forming the first transformer temperature prediction time-series information. This retained data is crucial for subsequent transformer oil chromatography monitoring and anomaly early warning. This prediction method is particularly suitable for scenarios where environmental temperature variations are significant and have a substantial impact on transformer temperature. For example, under extreme weather conditions, such as extreme heat or cold, the deviation between the environmental temperature and the reference temperature may be large. In such cases, using this data statistical strategy for temperature prediction can more accurately reflect the actual trend of transformer temperature changes, providing a more reliable temperature reference for subsequent oil chromatography monitoring.

[0024] When the deviation between the monitored ambient temperature and the reference ambient temperature remains within the ambient temperature deviation threshold, it indicates that the current ambient temperature is relatively stable and there are no significant fluctuations. In this case, to quickly and efficiently obtain transformer temperature prediction information, a temperature predictor bound to the transformer's tag number is used. This temperature predictor is specifically trained based on the operating parameters and historical data of a particular transformer, accurately reflecting the temperature change pattern of the transformer under different control parameters. In this process, the temperature predictor associated with the transformer tag number is first retrieved. This predictor has been pre-trained and optimized based on the transformer's control parameter time series matrix (containing various operating parameters of the transformer and their changes over time). Then, the current monitored ambient temperature and the transformer control parameter time series matrix are input into this temperature predictor. Based on this input information, the temperature predictor uses its internal algorithms and models to quickly calculate and output the second transformer temperature prediction time series information. This second transformer temperature prediction time series information represents the possible trend and range of transformer temperature changes under the current ambient temperature and control parameters. Because this prediction method directly utilizes a pre-trained temperature predictor, it can greatly improve prediction efficiency. Meanwhile, because the temperature predictor is trained based on the operating parameters and historical data of specific transformers, it can better adapt to the characteristics and operating environments of different transformers. For example, in spring and autumn, the ambient temperature is relatively stable. Although there may be some fluctuations, if these fluctuations remain within the ambient temperature deviation threshold, then this rapid prediction method can obtain transformer temperature prediction information without the need for complex data retrieval and statistical analysis. In this way, the real-time performance and efficiency of monitoring can be improved while ensuring prediction accuracy.

[0025] Subsequently, based on the obtained first or second transformer temperature prediction time series information, and combined with the service life of the transformer oil, the centroidal transformer oil chromatograms of similar healthy transformer oil sample groups are retrieved. The first and second transformer temperature prediction time series information represent the transformer temperature change trends and ranges obtained under different ambient temperature conditions using different prediction strategies, respectively. The service life of the transformer oil refers to the time the transformer oil has been used, reflecting its aging degree and possible chemical changes. Using this information, healthy sample groups with similar transformer oil conditions to the current transformer oil are retrieved from the database. The transformer oils in these sample groups are comparable to the current transformer oil in terms of temperature, service life, etc., therefore their centroidal transformer oil chromatograms can serve as reference standards. The centroidal transformer oil chromatograms represent the average or typical chromatographic characteristics of the transformer oils in these healthy sample groups, including the types and contents of various substances. Furthermore, a reference substance type list and a reference content range list are extracted from the centroidal transformer oil chromatograms. The reference substance type list lists the types of substances that should be present in healthy transformer oil, while the reference content range list gives the content range of these substances in healthy transformer oil. Finally, chromatographic monitoring is performed on the current transformer oil based on this baseline information. By comparing the types and contents of substances in the current transformer oil with the baseline substance type list and baseline content range list, the health status of the current transformer oil can be determined, and potential abnormalities or faults can be detected in a timely manner. This monitoring method is highly adaptable. For example, for a transformer that has been operating for a long time, its transformer oil may have a high degree of aging. In this case, a corresponding healthy sample group will be retrieved based on its service life, and the baseline substance type list and baseline content range list will be adjusted accordingly to ensure the accuracy of the monitoring. Similarly, for transformers operating at different ambient temperatures, appropriate healthy sample groups will be selected for reference based on temperature prediction time series information, thereby improving the targeting and effectiveness of the monitoring.

[0026] In a preferred embodiment, when the environmental temperature deviation vector between the monitored environmental temperature and the reference environmental temperature is greater than or equal to an environmental temperature deviation threshold, the process includes: collecting a first transformer temperature monitoring log, wherein the first transformer temperature monitoring log includes a transformer control parameter timing record matrix, environmental temperature record information, and first transformer temperature record timing information; collecting second transformer temperature record timing information constrained by the transformer control parameter timing record matrix and the reference environmental temperature; when the temperature timing deviation between the first transformer temperature record timing information and the second transformer temperature record timing information is greater than or equal to a temperature timing deviation threshold, storing the first environmental temperature deviation between the environmental temperature record information and the reference environmental temperature into a discrete environmental temperature deviation threshold set; when the number of discrete environmental temperature deviation thresholds is greater than or equal to a predefined number threshold, extracting the minimum value of the set of environmental temperature deviation thresholds and setting it as the environmental temperature deviation threshold.

[0027] Optionally, the configuration of the ambient temperature deviation threshold determines the decision-making of the temperature prediction strategy. In this scheme, when the ambient temperature deviation vector between the monitored ambient temperature and the reference ambient temperature is greater than or equal to the current ambient temperature deviation threshold, this threshold needs to be dynamically adjusted. The current ambient temperature deviation threshold is a preset value used in transformer oil chromatography monitoring to determine whether the deviation between the monitored ambient temperature and the reference ambient temperature is significant. It is determined based on factors such as the transformer's operating characteristics, historical data, and industry standards. Specifically, when the deviation vector (i.e., the difference or change between the two) between the monitored ambient temperature and the reference ambient temperature is greater than or equal to this preset ambient temperature deviation threshold, it is considered that the current ambient temperature's impact on the transformer has exceeded the normal range, thus triggering a more refined temperature prediction and monitoring strategy. Furthermore, the ambient temperature deviation threshold is not fixed but can be dynamically adjusted according to actual operating conditions. For example, by accumulating and analyzing historical data, this threshold can be automatically adjusted to better adapt to the actual operating environment of the transformer. For instance, assuming that the ambient temperature around a transformer is typically maintained between 20°C and 30°C during normal operation, to ensure the safe operation of the transformer, the ambient temperature deviation threshold can be set to 5°C. This means that when the ambient temperature exceeds 35℃ or falls below 15℃, the system is considered to have exceeded the normal range of influence on the transformer, triggering corresponding temperature prediction and monitoring strategies. However, if the transformer frequently operates in environments above 35℃ during the hot summer months, and historical data analysis shows that the transformer can still operate normally within this temperature range, the ambient temperature deviation threshold may be dynamically adjusted to a higher value, such as 7℃. In subsequent monitoring, the corresponding strategy will only be triggered when the ambient temperature exceeds 37℃ or falls below 13℃. This dynamic adjustment mechanism improves the system's adaptability and accuracy.

[0028] Specifically, the process begins by collecting the temperature monitoring log of the first transformer. This log details the timing of the transformer's control parameters, ambient temperature, and temperature changes, forming the basis for subsequent analysis. Next, using the transformer control parameter timing record matrix and a reference ambient temperature as constraints, the timing information of the second transformer's temperature records is collected. The purpose of this step is to obtain the expected temperature change trend of the transformer under the reference ambient temperature. Then, the temperature timing deviation between the first and second transformer temperature records is compared. If this deviation is greater than or equal to a preset temperature timing deviation threshold, it indicates that the current ambient temperature's impact on the transformer temperature has exceeded the normal range. In this case, the current ambient temperature record information and the first ambient temperature deviation from the reference ambient temperature are stored in a discrete ambient temperature deviation threshold set. As operation continues, the elements in the discrete ambient temperature deviation threshold set accumulate. When the number of elements in this set is greater than or equal to a predefined threshold, the minimum value of all ambient temperature deviation thresholds in the set is extracted and set as the new ambient temperature deviation threshold. Essentially, this step dynamically adjusts the ambient temperature deviation threshold by analyzing historical data to better reflect the actual operating environment of the transformer. For example, during the hot summer months, transformers may frequently operate in high-temperature environments, leading to significant deviations between the ambient temperature and the reference ambient temperature. In such cases, by continuously accumulating ambient temperature deviation data and dynamically adjusting the ambient temperature deviation threshold through the aforementioned process, it becomes possible to more accurately determine when a more refined temperature prediction strategy is needed in subsequent monitoring. This method of dynamically configuring the ambient temperature deviation threshold improves the adaptability and accuracy of the transformer oil chromatography monitoring system.

[0029] In a preferred embodiment, when the temperature time-series deviation between the first transformer temperature recording time-series information and the second transformer temperature recording time-series information is greater than or equal to a temperature time-series deviation threshold, the method includes: calculating the transformer temperature deviation magnitude time-series information between the first transformer temperature recording time-series information and the second transformer temperature recording time-series information; statistically analyzing the proportion of times in the transformer temperature deviation magnitude time-series information where the temperature deviation magnitude is greater than or equal to the temperature deviation magnitude threshold, and setting this as a first deviation parameter; calculating the dynamic time warping distance between the first transformer temperature recording time-series information and the second transformer temperature recording time-series information, and setting this as a second deviation parameter; the temperature time-series deviation threshold includes a first deviation threshold and a second deviation threshold; when the first deviation parameter is greater than or equal to the first deviation threshold and the second deviation parameter is greater than or equal to the second deviation threshold, the temperature time-series deviation is considered to be greater than or equal to the temperature time-series deviation threshold.

[0030] Specifically, when it is necessary to determine whether the temperature time series deviation between the first transformer temperature record time series information and the second transformer temperature record time series information is greater than or equal to the temperature time series deviation threshold, the transformer temperature deviation magnitude time series information between these two time series information is first calculated. That is, the temperature values ​​in the two time series information are compared point by point, and the deviation magnitude between them is calculated to form a new time series information. The deviation magnitude time series information reflects the specific temperature difference between the two time series information. Then, the proportion of times in this deviation magnitude time series information where the temperature deviation magnitude is greater than or equal to the preset temperature deviation magnitude threshold is counted. This proportion is set as the first deviation parameter. The first deviation parameter actually reflects the frequency of significant temperature deviations between the two time series information. At the same time, the dynamic time warping distance between the first transformer temperature record time series information and the second transformer temperature record time series information is calculated. This distance is set as the second deviation parameter. The dynamic time warping distance is a method to measure the similarity between two time series information. It takes into account the time alignment problem in the time series information, and therefore can more accurately reflect the differences in shape and trend between the two time series information.

[0031] The temperature time-series deviation threshold can be understood as a composite threshold, comprising a first deviation threshold and a second deviation threshold. Only when both the first deviation parameter and the second deviation parameter are greater than or equal to the first deviation threshold are the temperature time-series deviation considered greater than or equal to the temperature time-series deviation threshold. For example, suppose in a monitoring session, the time-series information of the first transformer temperature record shows that the transformer temperature remained consistently high for a period of time, while the time-series information of the second transformer temperature record shows that the transformer temperature fluctuated within the normal range. By calculating the deviation magnitude time-series information and the dynamic time warping distance, it is found that both the first and second deviation parameters exceed the preset thresholds. This means that there is a significant difference in temperature between the two time-series information, and this difference is not accidental but persistent. Therefore, the temperature time-series deviation is determined to be greater than or equal to the temperature time-series deviation threshold, triggering the corresponding early warning or processing mechanism. This refined analytical method improves the accuracy and reliability of transformer oil chromatography monitoring.

[0032] In a preferred embodiment, retrieving the first transformer temperature prediction time series information that satisfies the transformer control parameter time series matrix and the environmental monitoring temperature includes: retrieving a set of transformer temperature monitoring time series information that satisfies the transformer control parameter time series matrix and the environmental monitoring temperature; performing cluster analysis on the set of transformer temperature monitoring time series information to obtain multi-cluster transformer temperature monitoring time series information; deleting cluster data where the number of transformer temperature monitoring time series information within a cluster is less than or equal to a cluster quantity threshold from the set of transformer temperature monitoring time series information to obtain retained transformer temperature monitoring time series information; and statistically analyzing the temporal temperature distribution interval of the retained transformer temperature monitoring time series information, which is set as the first transformer temperature prediction time series information.

[0033] In detail, the prediction process for the first transformer temperature prediction time series information aims to accurately predict the temperature change trend of the transformer based on the transformer's control parameter time series matrix and the environmental monitoring temperature. Specifically, it retrieves a set of transformer temperature monitoring time series information that meets the specific transformer control parameter time series matrix and environmental monitoring temperature. This set contains a large amount of historical monitoring data, reflecting the temperature changes of the transformer under different control parameters and environmental temperatures. Cluster analysis is then performed on this set. Cluster analysis is an unsupervised learning method that can divide data points into multiple clusters based on similarity. In this scenario, the purpose of cluster analysis is to group similar transformer temperature monitoring time series information into one category for subsequent analysis. Multiple clusters of transformer temperature monitoring time series information are obtained through cluster analysis. These clusters are then further filtered. Specifically, clusters with a number of transformer temperature monitoring time series information points less than or equal to a cluster size threshold are removed from the set. This step aims to eliminate clusters that may be formed due to abnormal or noisy data, retaining clusters containing a sufficient amount of valid data. After filtering, the retained transformer temperature monitoring time series information is obtained. Finally, the temporal temperature distribution range of these retained transformer temperature monitoring time series information is statistically analyzed. This distribution range represents the potential temperature variation range of the transformer under current control parameters and ambient temperature. This distribution range is set as the primary transformer temperature prediction time series information for subsequent transformer oil chromatography monitoring and anomaly early warning. For example, suppose that in a prediction, a large amount of temperature monitoring time series information for a certain transformer is retrieved. Through cluster analysis, it is found that this data can be divided into several different clusters, each representing the temperature change pattern of the transformer under different operating conditions. After filtering, clusters containing sufficient valid data are retained, and the temperature distribution range within these clusters at the same time is statistically analyzed. Finally, based on this distribution range, the future temperature change trend of the transformer is predicted, providing an important basis for subsequent monitoring and early warning.

[0034] In a preferred embodiment, a second transformer temperature prediction time series information is obtained by processing the transformer control parameter time series matrix through a temperature predictor bound to the transformer tag number. This includes: retrieving the annual operating environment temperature record time series information of the transformer tag number; performing adjacent time domain temperature hierarchical clustering analysis on the annual operating environment temperature record time series information to obtain a first time domain reference environment temperature up to the Nth time domain reference environment temperature; traversing the first time domain reference environment temperature up to the Nth time domain reference environment temperature to build a first temperature predictor up to the Nth temperature predictor bound to the transformer tag number; configuring the reference environment temperature according to the monitoring time, selectively scheduling the first temperature predictor up to the Nth temperature predictor, processing the transformer control parameter time series matrix, and obtaining the second transformer temperature prediction time series information.

[0035] Preferably, the process of processing the transformer control parameter time series matrix to obtain the second transformer temperature prediction time series information using a temperature predictor bound to the transformer tag number is a process that combines historical data analysis and real-time prediction strategies. Specifically, the annual operating environment temperature record time series information for a specific transformer tag number is retrieved. This information records the environmental temperature conditions of the transformer at different points in time over the past year. Adjacent time domain temperature hierarchical clustering analysis is performed on these annual operating environment temperature record time series information. The purpose of this step is to identify the similarities and differences in environmental temperatures across different time periods, thereby dividing the environment into multiple time domain reference environmental temperatures, such as the first time domain reference environmental temperature, the second time domain reference environmental temperature, and so on up to the Nth time domain reference environmental temperature. These reference environmental temperatures represent the typical or average state of environmental temperature across different time periods.

[0036] Next, the time-domain reference ambient temperatures are iterated, and a temperature predictor bound to the transformer tag number is built for each reference ambient temperature, i.e., the first temperature predictor to the Nth temperature predictor. These predictors are trained based on historical operating data of the transformer under specific ambient temperatures and can accurately predict the temperature change trend of the transformer under those temperatures. In the actual prediction process, the corresponding reference ambient temperature is configured according to the current monitoring time. This means selecting the time-domain reference ambient temperature that best matches the current time, or weighting multiple time-domain reference ambient temperatures to more accurately reflect the current environmental conditions. Finally, the first to Nth temperature predictors are selectively scheduled according to the configured reference ambient temperatures. That is, one or more predictors best suited to the current environmental conditions are selected to process the transformer control parameter timing matrix, thereby obtaining the second transformer temperature prediction timing information. This prediction timing information represents the possible trend and range of transformer temperature change under the current environmental conditions.

[0037] For example, suppose a transformer experiences consistently high summer temperatures throughout the past year, and these temperatures occur frequently. In adjacent time-domain temperature hierarchical clustering analysis, these high summer temperatures are likely to be identified as a significant time-domain baseline temperature. The temperature predictor will be specifically trained and optimized for this high-temperature environment. When summer arrives and the current ambient temperature matches this high-temperature baseline, the appropriate temperature predictor will be selected to process the transformer's control parameter time-series matrix, thus more accurately predicting the transformer's temperature change trend under high-temperature conditions. This method of selecting a temperature that occurs most frequently each year as the baseline temperature improves the accuracy and adaptability of the prediction.

[0038] In a preferred embodiment, traversing the first time-domain reference ambient temperature up to the Nth time-domain reference ambient temperature, and constructing a first temperature predictor bound to the transformer tag number up to the Nth temperature predictor, includes: collecting multiple sets of data under the constraints of the first time-domain reference ambient temperature and the transformer model; wherein any set of the multiple sets of data includes a transformer control parameter timing record matrix and transformer temperature recording timing information; using the transformer temperature recording timing information as supervision and the transformer control parameter timing record matrix as input, training the first temperature predictor through machine learning; until training the Nth temperature predictor.

[0039] Furthermore, in constructing the temperature predictor bound to the transformer reference number, the process iterates from the first time-domain reference ambient temperature to the Nth time-domain reference ambient temperature, building a corresponding temperature predictor for each reference ambient temperature. Specifically, for each time-domain reference ambient temperature, multiple sets of data are collected, using the reference ambient temperature and transformer model as constraints. Each set of these data contains a transformer control parameter timing record matrix and transformer temperature recording timing information, representing the transformer's operation under different control parameters and the corresponding temperature changes, respectively. Next, using the transformer temperature recording timing information as a supervisory signal and the transformer control parameter timing record matrix as input, a machine learning algorithm is used to train the temperature predictor. During training, the machine learning algorithm learns the mapping relationship between control parameters and temperature, thus enabling it to predict the corresponding transformer temperature recording timing information based on the input control parameter timing record matrix. This process is repeated until the Nth temperature predictor corresponding to the Nth time-domain reference ambient temperature is trained. Each temperature predictor is trained for a specific reference ambient temperature and transformer model, therefore they can more accurately predict the temperature change trend of the transformer under that reference ambient temperature. For example, assuming a transformer operates in a high-temperature environment during summer, the summer high temperature would be used as the first time-domain reference ambient temperature, and multiple sets of data for the transformer under these conditions would be collected. Then, using this data as a training set, a machine learning algorithm would be used to train a first temperature predictor for the summer high-temperature environment. Similarly, corresponding data would be collected for other time-domain reference ambient temperatures, and corresponding temperature predictors would be trained. In this way, when the transformer operates under different ambient temperatures, the most suitable temperature predictor for the current ambient temperature can be selected for prediction, thereby improving the accuracy and adaptability of the prediction.

[0040] In a preferred embodiment, based on the first transformer temperature prediction time series information or the second transformer temperature prediction time series information, and combined with the transformer oil service life, the centroidal transformer oil chromatograms of similar healthy transformer oil sample groups are retrieved to obtain a list of reference substance types and a list of reference content ranges. This includes: retrieving a primary group of similar healthy transformer oil sample groups using the first transformer temperature prediction time series information or the second transformer temperature prediction time series information and the transformer oil service life as constraints; retrieving a secondary group of similar healthy transformer oil sample groups using the primary transformer oil service life of the primary group of similar healthy transformer oil sample groups and the first transformer temperature prediction time series information as constraints; and statistically analyzing the centroidal transformer oil chromatograms of the primary and secondary similar healthy transformer oil sample groups to obtain a list of reference substance types and a list of reference content ranges.

[0041] Specifically, to obtain accurate lists of reference material types and reference content ranges, a multi-level indexing strategy for sample expansion was adopted. First, using the temperature prediction time series information of either the first or second transformer, along with the service life of the transformer oil, as constraints, a primary-level healthy sample group of similar transformer oils was retrieved. The transformer oils in these sample groups are similar to the current transformer oil in terms of temperature prediction time series information and service life, and therefore can serve as initial comparison objects. Next, using the service life of the primary-level transformer oils in the primary-level healthy sample group and the temperature prediction time series information of the first transformer as further constraints, a secondary-level healthy sample group of similar transformer oils was retrieved. The purpose of this step is to screen samples at a more refined level, ensuring that the secondary sample group has a higher similarity to the current transformer oil in terms of service life and temperature prediction time series information.

[0042] After obtaining primary and secondary healthy transformer oil sample groups, the centroidal transformer oil chromatograms of these sample groups are statistically analyzed. Centroidal transformer oil chromatograms represent the average or typical chromatographic characteristics of the transformer oils in these sample groups, including the types and contents of various substances. Statistical analysis of these centroidal transformer oil chromatograms allows for the extraction of a list of reference substance types and a list of reference content ranges. This multi-level indexing strategy for expanding the sample improves the accuracy and efficiency of sample screening. For example, assuming a transformer oil has a long service life and its temperature prediction time series information indicates that it operates in a high-temperature environment, all primary healthy transformer oil sample groups with similar service lives and operating in high-temperature environments will first be retrieved. Then, from these primary sample groups, secondary healthy transformer oil sample groups with service lives and temperature prediction time series information more closely related to the current transformer oil will be further screened. Finally, by statistically analyzing the centroidal transformer oil chromatograms of these secondary sample groups, a more accurate list of reference substance types and a list of reference content ranges can be obtained.

[0043] The transformer oil chromatography monitoring method provided in this embodiment of the invention has at least the following technical effects: 1. By monitoring the deviation vector between the monitored ambient temperature and the reference ambient temperature in real time and dynamically adjusting the ambient temperature deviation threshold, the system can more flexibly adapt to temperature changes under different operating conditions. When the ambient temperature deviation is large, a more refined temperature prediction and monitoring strategy is triggered, thereby improving the accuracy and reliability of monitoring. This dynamic adjustment mechanism helps reduce false alarms and missed alarms caused by changes in ambient temperature, enhancing the intelligence level of transformer oil chromatographic monitoring.

[0044] 2. By using a temperature predictor bound to the transformer tag number and combining it with the timing matrix of transformer control parameters, accurate prediction of transformer temperature is achieved. Specifically, through hierarchical clustering analysis of adjacent time-domain temperatures, multiple time-domain reference ambient temperatures are obtained, and a corresponding temperature predictor is built for each reference ambient temperature. In practical applications, the most suitable temperature predictor is selectively scheduled based on the ambient temperature at the monitoring time, thereby improving the accuracy and adaptability of temperature prediction. This combination of multi-level temperature predictors and selective scheduling strategies can better cope with complex and changing operating environments.

[0045] 3. A multi-level indexing strategy was employed to expand the sample when obtaining the list of reference substance types and the list of reference content ranges. First, using transformer temperature prediction time series information and transformer oil service life as constraints, a primary group of healthy transformer oil samples of the same type was retrieved. Then, further using the service life and temperature prediction time series information of the primary sample group as constraints, a secondary group of healthy transformer oil samples of the same type was retrieved. By statistically analyzing the centroid transformer oil chromatograms of these sample groups, the list of reference substance types and the list of reference content ranges could be extracted more accurately. This multi-level indexing method helps to broaden the sample range, improve the representativeness and accuracy of the samples, thereby enhancing the precision and reliability of transformer oil chromatographic monitoring. Example

[0046] like Figure 2 As shown, based on the same inventive concept as the transformer oil chromatography monitoring method provided in Embodiment 1, this embodiment of the invention also provides a transformer oil chromatography monitoring system, the system comprising: Information retrieval module 11 is used to retrieve first transformer temperature prediction timing information that satisfies the transformer control parameter timing matrix and the environmental monitoring temperature when the environmental temperature deviation vector between the environmental monitoring temperature and the reference environmental temperature is greater than or equal to the environmental temperature deviation threshold.

[0047] The data processing module 12 is used to process the timing matrix of the transformer control parameters through a temperature predictor bound to the transformer tag number when the ambient temperature deviation vector between the monitored ambient temperature and the reference ambient temperature is less than the ambient temperature deviation threshold, thereby obtaining the timing information of the second transformer temperature prediction.

[0048] The monitoring execution module 13 is used to retrieve the centroid transformer oil chromatograms of similar transformer oil healthy sample groups based on the first transformer temperature prediction time series information or the second transformer temperature prediction time series information, combined with the transformer oil service time, to obtain a list of reference substance types and a list of reference content ranges, and to perform transformer oil chromatographic monitoring.

[0049] Furthermore, the information retrieval module 11 is also used to perform the following steps: A temperature monitoring log for the first transformer is collected, comprising a transformer control parameter timing record matrix, ambient temperature record information, and a timing information of the first transformer temperature record. A timing information of the second transformer temperature record is collected, constrained by the transformer control parameter timing record matrix and the reference ambient temperature. When the temperature timing deviation between the first and second transformer temperature record timing information is greater than or equal to a temperature timing deviation threshold, the first ambient temperature deviation between the ambient temperature record information and the reference ambient temperature is stored in a discrete ambient temperature deviation threshold set. When the number of values ​​in the discrete ambient temperature deviation threshold set is greater than or equal to a predefined threshold value, the minimum value of the set's ambient temperature deviation thresholds is extracted and set as the ambient temperature deviation threshold.

[0050] Furthermore, the information retrieval module 11 is also used to perform the following steps: Calculate the transformer temperature deviation magnitude time-series information between the first transformer temperature record time-series information and the second transformer temperature record time-series information; statistically analyze the proportion of times in the transformer temperature deviation magnitude time-series information where the temperature deviation magnitude is greater than or equal to a temperature deviation magnitude threshold, and set this as a first deviation parameter; calculate the dynamic time warping distance between the first transformer temperature record time-series information and the second transformer temperature record time-series information, and set this as a second deviation parameter; the temperature time-series deviation threshold includes a first deviation threshold and a second deviation threshold; when the first deviation parameter is greater than or equal to the first deviation threshold and the second deviation parameter is greater than or equal to the second deviation threshold, the temperature time-series deviation is considered to be greater than or equal to the temperature time-series deviation threshold.

[0051] Furthermore, the information retrieval module 11 is also used to perform the following steps: Retrieve a set of transformer temperature monitoring time-series information that satisfies the transformer control parameter time-series matrix and the environmental monitoring temperature; perform cluster analysis on the set of transformer temperature monitoring time-series information to obtain multi-cluster transformer temperature monitoring time-series information; delete cluster data where the number of transformer temperature monitoring time-series information within a cluster is less than or equal to a cluster quantity threshold from the set of transformer temperature monitoring time-series information to obtain retained transformer temperature monitoring time-series information; statistically analyze the temporal temperature distribution interval of the retained transformer temperature monitoring time-series information and set it as the first transformer temperature prediction time-series information.

[0052] Furthermore, the data processing module 12 is also used to perform the following steps: Retrieve the annual operating ambient temperature record time series information for the transformer tag number; perform adjacent time domain temperature hierarchical clustering analysis on the annual operating ambient temperature record time series information to obtain the first time domain reference ambient temperature up to the Nth time domain reference ambient temperature; traverse the first time domain reference ambient temperature up to the Nth time domain reference ambient temperature, and build the first temperature predictor up to the Nth temperature predictor bound to the transformer tag number; configure the reference ambient temperature according to the monitoring time, selectively schedule the first temperature predictor up to the Nth temperature predictor, process the transformer control parameter time series matrix, and obtain the second transformer temperature prediction time series information.

[0053] Furthermore, the data processing module 12 is also used to perform the following steps: Multiple sets of data are collected, constrained by the first time-domain reference ambient temperature and the transformer model. Each set of data includes a transformer control parameter timing record matrix and a transformer temperature recording timing information. The transformer temperature recording timing information is used as supervision, and the transformer control parameter timing record matrix is ​​used as input. The first temperature predictor is trained through machine learning until the Nth temperature predictor is trained.

[0054] Furthermore, the monitoring execution module 13 is also used to perform the following steps: Using the first transformer temperature prediction time series information or the second transformer temperature prediction time series information, and the service life of the transformer oil as constraints, a primary-level healthy sample group of similar transformer oils is retrieved; using the service life of the primary-level transformer oils in the primary-level healthy sample group of similar transformer oils and the first transformer temperature prediction time series information as constraints, a secondary-level healthy sample group of similar transformer oils is retrieved; the centroidal transformer oil chromatography of the primary-level and secondary-level healthy sample groups of similar transformer oils is statistically analyzed to obtain a list of reference substance types and a list of reference content ranges.

[0055] Through the foregoing detailed description of a transformer oil chromatographic monitoring method, those skilled in the art can clearly understand the transformer oil chromatographic monitoring system described in this embodiment. As the system disclosed in the embodiment corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section description. Example

[0056] Please see Figure 3 , Figure 3 This is a schematic diagram of an embodiment of a smart terminal provided by the present invention. For example... Figure 3As shown, this embodiment of the invention provides a smart terminal 500, including a memory 510, a processor 520, and a first computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the first computer program 511, it implements a transformer oil chromatography monitoring method as described in Embodiment 1. Example

[0057] Please see Figure 4 , Figure 4 This is a schematic diagram illustrating an embodiment of a computer-readable storage medium provided by an embodiment of the present invention. For example... Figure 4 As shown, this embodiment provides a computer-readable storage medium 600, on which a second computer program 611 is stored. When the second computer program 611 is executed by a processor, it implements a transformer oil chromatography monitoring method as described in Embodiment 1.

[0058] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0059] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0060] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0061] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0062] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0063] Although preferred embodiments of the invention have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0064] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for chromatographic monitoring of transformer oil, characterized in that, include: When the ambient temperature deviation vector between the environmental monitoring temperature and the reference ambient temperature is greater than or equal to the ambient temperature deviation threshold, the first transformer temperature prediction timing information that satisfies the transformer control parameter timing matrix and the environmental monitoring temperature is retrieved. When the ambient temperature deviation vector between the monitored ambient temperature and the reference ambient temperature is less than the ambient temperature deviation threshold, the temperature predictor bound to the transformer tag number processes the timing matrix of the transformer control parameters to obtain the timing information of the second transformer temperature prediction. Based on the first transformer temperature prediction time series information or the second transformer temperature prediction time series information, combined with the transformer oil service life, the centroid transformer oil chromatography of similar healthy transformer oil sample groups is retrieved to obtain a list of reference substance types and a list of reference content ranges, and then transformer oil chromatography monitoring is performed.

2. The method as described in claim 1, characterized in that, When the environmental temperature deviation vector between the monitored environmental temperature and the reference environmental temperature is greater than or equal to the environmental temperature deviation threshold, including: Collect the temperature monitoring log of the first transformer, wherein the temperature monitoring log of the first transformer includes a timing record matrix of transformer control parameters, ambient temperature record information and timing information of the first transformer temperature record; Using the transformer control parameter timing record matrix and the reference ambient temperature as constraints, the timing information of the second transformer temperature record is collected; When the temperature timing deviation between the first transformer temperature recording timing information and the second transformer temperature recording timing information is greater than or equal to the temperature timing deviation threshold, the first ambient temperature deviation between the ambient temperature recording information and the reference ambient temperature is stored in the discrete ambient temperature deviation threshold set. When the number of discrete environmental temperature deviation thresholds is greater than or equal to a predefined threshold, the minimum value of the set of environmental temperature deviation thresholds is extracted and set as the environmental temperature deviation threshold.

3. The method as described in claim 2, characterized in that, When the temperature timing deviation between the first transformer temperature recording timing information and the second transformer temperature recording timing information is greater than or equal to the temperature timing deviation threshold, including: Calculate the transformer temperature deviation modulus timing information between the timing information of the first transformer temperature record and the timing information of the second transformer temperature record; The percentage of times when the temperature deviation magnitude is greater than or equal to the temperature deviation magnitude threshold in the time sequence information of the transformer temperature deviation magnitude is statistically analyzed and set as the first deviation parameter. Calculate the dynamic time warping distance between the timing information of the first transformer temperature record and the timing information of the second transformer temperature record, and set it as the second deviation parameter; The temperature time-series deviation threshold includes a first deviation threshold and a second deviation threshold; When the first deviation parameter is greater than or equal to the first deviation threshold, and the second deviation parameter is greater than or equal to the second deviation threshold, the temperature time series deviation is considered to be greater than or equal to the temperature time series deviation threshold.

4. The method as described in claim 1, characterized in that, Retrieve the first transformer temperature prediction timing information that satisfies the transformer control parameter timing matrix and the environmental monitoring temperature, including: Retrieve a set of transformer temperature monitoring timing information that satisfies the transformer control parameter timing matrix and the environmental monitoring temperature; Cluster analysis is performed on the transformer temperature monitoring time series information set to obtain multi-cluster transformer temperature monitoring time series information; Clusters whose number of transformer temperature monitoring time sequence information within a cluster is less than or equal to the cluster quantity threshold are deleted from the transformer temperature monitoring time sequence information set to obtain retained transformer temperature monitoring time sequence information. The temperature distribution interval at the same time as the retained transformer temperature monitoring time series information is statistically analyzed and set as the first transformer temperature prediction time series information.

5. The method as described in claim 1, characterized in that, By processing the timing matrix of the transformer control parameters through a temperature predictor bound to the transformer tag number, the timing information for predicting the temperature of the second transformer is obtained, including: Retrieve the time-series information of the annual operating environment temperature records for the transformer tag number; Perform adjacent time-domain temperature hierarchical clustering analysis on the time-series information of the annual operating environment temperature records to obtain the first time-domain reference environment temperature up to the Nth time-domain reference environment temperature. Iterate through the first time-domain reference ambient temperature up to the Nth time-domain reference ambient temperature, and build a first temperature predictor bound to the transformer reference number up to the Nth temperature predictor. Based on the monitoring time, the reference ambient temperature is configured, and the first temperature predictor up to the Nth temperature predictor is selectively scheduled. The transformer control parameter timing matrix is ​​processed to obtain the second transformer temperature prediction timing information.

6. The method as described in claim 5, characterized in that, Traversing the first time-domain reference ambient temperature up to the Nth time-domain reference ambient temperature, constructing a first temperature predictor bound to the transformer reference number up to the Nth temperature predictor, including: Multiple sets of data were collected, using the first time-domain reference ambient temperature and transformer model as constraints. Each of the multiple sets of data includes a transformer control parameter timing record matrix and a transformer temperature recording timing information; Using the transformer temperature recording time series information as supervision and the transformer control parameter time series recording matrix as input, the first temperature predictor is trained through machine learning. This continues until the Nth temperature predictor is trained.

7. The method as described in claim 1, characterized in that, Based on the predicted time series information of the first or second transformer temperature, and combined with the service life of the transformer oil, the centroid transformer oil chromatograms of similar healthy transformer oil sample groups are retrieved to obtain a list of reference substance types and a list of reference content ranges, including: Using the first transformer temperature prediction time series information or the second transformer temperature prediction time series information, and the transformer oil service life as constraints, a first-level similar transformer oil health sample group is retrieved. Using the service life of the first-level transformer oil in the first-level similar transformer oil health sample group and the temperature prediction time series information of the first transformer as constraints, a second-level similar transformer oil health sample group is retrieved. By statistically analyzing the centroidal transformer oil chromatograms of the first-class and second-class healthy transformer oil sample groups, a list of reference substance types and a list of reference content ranges were obtained.

8. A transformer oil chromatography monitoring system, characterized in that, For implementing the transformer oil chromatographic monitoring method according to any one of claims 1-7, the system comprises: The information retrieval module is used to retrieve the first transformer temperature prediction timing information that satisfies the transformer control parameter timing matrix and the environmental monitoring temperature when the environmental temperature deviation vector between the environmental monitoring temperature and the reference environmental temperature is greater than or equal to the environmental temperature deviation threshold. The data processing module is used to process the timing matrix of transformer control parameters through a temperature predictor bound to the transformer tag number when the ambient temperature deviation vector between the monitored ambient temperature and the reference ambient temperature is less than the ambient temperature deviation threshold, and obtain the timing information of the second transformer temperature prediction. The monitoring execution module is used to retrieve the centroid transformer oil chromatograms of similar healthy transformer oil sample groups based on the first transformer temperature prediction time series information or the second transformer temperature prediction time series information, combined with the transformer oil service life, to obtain a list of reference substance types and a list of reference content ranges, and to perform transformer oil chromatographic monitoring.

9. A smart terminal, characterized in that, include: Memory, used to store computer software programs; A processor is configured to read and execute the computer software program, thereby implementing the transformer oil chromatographic monitoring method according to any one of claims 1-7.

10. A non-transitory computer-readable storage medium, characterized in that, The storage medium stores a computer software program, which, when executed by a processor, implements a transformer oil chromatographic monitoring method as described in any one of claims 1-7.

Images