A transformer performance evaluation method based on data analysis and its application

Abstract

The present invention relates to the technical field of transformer performance evaluation, and specifically discloses a transformer performance evaluation method and application based on data analysis. The method comprises the following steps: grouping transformers based on changes in environmental factors, marking transformers in the group with the same transformation method as target transformers; obtaining the time point t1 when the transformation of the target transformer begins and the time point t2 when the transformation is completed, and setting a plurality of first time points and second time points based on time points t1 and t2; obtaining a first score and a second score, classifying the target transformers based on the first score and the second score, determining a representative transformer in a single category, and evaluating the feasibility of the transformation of the target transformer based on the first score and the second score of the representative transformer. The present invention can evaluate the feasibility of transformer transformation.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer performance evaluation, and in particular to a transformer performance evaluation method based on data analysis and its application. Background Art

[0002] A transformer is an electrical device that uses the principle of electromagnetic induction to convert AC voltage and current proportionally. It consists primarily of two parts: an iron core and windings. When AC current flows through the primary winding, a changing magnetic flux is generated, which is conducted through the iron core to the secondary winding, inducing a corresponding voltage across it, thereby increasing or decreasing the voltage.

[0003] To meet the needs of different voltage regulation and operating conditions, the transformer can be modified. Common modification methods include: 1) Rewinding the original no-load tap winding or adding on-load tap taps, determining the number of turns, wire diameter and structure of each tap according to calculation, and properly leading out the tap leads to achieve online voltage regulation and fine-tune the output voltage without shutting down the machine; 2) Opening holes in the oil tank side panel or oil pillow, welding and sealing the on-load tap changer, and at the same time processing the box size and sealing structure as required, and installing a special oil tank to provide reliable mechanical support and a good cooling and sealing environment for the on-load voltage regulation device, thereby extending the life of the switch; 3) According to the voltage level and rated capacity of the transformer, select an on-load tap changer model with a step difference of ≤2.5% and an appropriate rated current, and perform insulation and voltage withstand tests on the entire system to ensure voltage regulation accuracy and electrical safety, so that the modified transformer can operate stably and reliably under various loads and operating conditions.

[0004] However, transformer retrofits are not always feasible. Improper winding turns or wire diameter design can cause resonance, localized core overheating, and even increased no-load losses. Poor sealing during oil tank machining or welding can lead to oil leakage or moisture absorption, degrading insulation performance and increasing the risk of breakdown. Misselected or inaccurately installed tap changers can cause poor contact and overheating, leading to tap failure and even damage to the switch structure. These conditions can cause previously high-performing transformers to experience a range of operational issues, including increased noise, reduced efficiency, and elevated temperature rise. In severe cases, these issues can even shorten the lifespan of the equipment. Therefore, assessing the feasibility of transformer retrofits has become a pressing issue. Summary of the Invention

[0005] The purpose of the present invention is to provide a transformer performance evaluation method and application based on data analysis to solve the above technical problems.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A transformer performance evaluation method based on data analysis includes the following steps:

[0008] Collect environmental factors around the transformers, group the transformers based on changes in these environmental factors, and mark transformers in the same group with the same modification method as target transformers;

[0009] Obtaining a time point t1 at which the target transformer starts transformation and a time point t2 at which the transformation is completed, setting a first period and a second period based on the time point t1 and the time point t2, and setting a plurality of first time points and second time points within the first period and the second period, respectively;

[0010] Performance scores of the target transformer are obtained at a first time point and a second time point, respectively recorded as a first score and a second score. The target transformer is classified based on the first score and the second score. For a single classification, a representative transformer is determined based on the first score and the second score of the target transformer in the classification. The feasibility of transformation of the target transformer is evaluated based on the first score and the second score of the representative transformer.

[0011] Preferably, grouping the transformers comprises:

[0012] Determine target factors among environmental factors whose impact on transformer performance exceeds a preset value based on deep learning technology, preprocess the target factors, including cleaning, denoising, and normalization, and fill in missing values ​​based on predicted values ​​from a machine learning algorithm;

[0013] Construct an n+1 dimensional coordinate system, where n represents the type of target factor, and generate coordinate points (A1, A2, ..., A n , S), A n represents the target factor after the nth preprocessing, and S represents the time point of collecting environmental factors;

[0014] The coordinate points are fitted to obtain a fitting curve. In the same group, the similarity between the fitting curves corresponding to any two transformers is greater than a preset first similarity threshold.

[0015] Preferably, setting the first time point and the second time point includes:

[0016] Set the first period [t1-T, t1] and the second period [t2, t2+T], where T represents the preset duration;

[0017] Starting from the starting points of the first cycle and the second cycle respectively, a number of time points are selected within the first cycle and the second cycle at preset time intervals, and are recorded as the first time point and the second time point respectively.

[0018] Preferably, obtaining the performance score of the target transformer includes:

[0019] Collecting operating data of the target transformer and performing a first processing on the operating data, the first processing including cleaning, denoising and normalization processing;

[0020] Using signal processing and machine learning algorithms, extracting influencing parameters from the operating data after the first processing, wherein the influencing parameters are characteristic parameters that affect the performance state of the transformer;

[0021] A transformer performance evaluation model is constructed based on the influencing parameters, the current influencing parameters are input into the transformer performance evaluation model, and the performance score of the target transformer is output.

[0022] Preferably, determining the representative transformer comprises:

[0023] Draw a curve A1 showing changes in the first score over time and a curve A2 showing changes in the second score over time;

[0024] In the same category, the first similarity corresponding to any two target transformers is greater than the preset second similarity threshold and the second similarity corresponding to any two target transformers is greater than the preset second similarity threshold. The first similarity represents the similarity between curves A1, and the second similarity represents the similarity between curves A2;

[0025] Calculate the selected value of target transformer i , p1 j and p2 j denote the first similarity and the second similarity between target transformer i and target transformer j respectively, and m denotes the total number of target transformers in the classification except target transformer i;

[0026] The target transformer corresponding to the maximum selection value is taken as the representative transformer.

[0027] Preferably, the feasibility assessment of the transformation of the target transformer includes:

[0028] Draw a curve d1 and a curve d2 representing the change of the first score of the transformer over time, respectively, and take the time points corresponding to the starting points of the curves d1 and d2 as the origins to obtain new curves, which are recorded as curve D1 and curve D2 respectively;

[0029] Get the curve D3 = D2 - D1, count the proportion X of the domain of the part below the x-axis on the curve D3 to the domain of the curve D3, and calculate the evaluation score , N represents the number of intersections of the curve D3 and the x-axis.

[0030] Preferably, the evaluation of the feasibility of the transformation of the target transformer further includes:

[0031] Set the evaluation score threshold K1;

[0032] If the evaluation score K≥K1, it is determined that the transformation feasibility of all target transformers in the classification is poor; otherwise, it is determined that the transformation feasibility of all target transformers in the classification is good.

[0033] An application of the transformer performance evaluation method based on data analysis in an operating transformer adopts a remote central computer to collect transformer operating data through SCADA, and applies the data to the evaluation of transformer transformation feasibility.

[0034] The beneficial effects of the present invention are as follows:

[0035] 1) The present invention first performs high-dimensional fitting on the collected environmental factor curves and automatically groups transformers based on curve similarity, placing equipment in the same group under a nearly identical environmental background, thereby shielding against the impact of environmental factors. This is because even if all operating data are identical, different performance evaluation results may result from different environmental factors. Subsequently, the performance score trends before and after the transformation are compared only within the same group, completely isolating the masking and amplification effects of environmental differences on the results, ensuring that the analysis conclusions truly reflect the impact of the transformation measures themselves, such as winding design, tank sealing, and tap changer selection. With this mechanism, operation and maintenance personnel can obtain a more objective, repeatable, and traceable evaluation basis, thus avoiding the risk of misjudging the feasibility of the transformation due to environmental factors at the source.

[0036] 2) This method selects the most representative equipment within each category. By performing in-depth curve differentiation and trend analysis on these representative transformers, the retrofit effects for all similar equipment can be inferred. This strategy significantly reduces the number of data cleaning, feature extraction, and curve comparisons, shortening the overall evaluation cycle and avoiding the system load caused by repeated calculations. Furthermore, the operations and maintenance team can quickly complete the feasibility screening of large-scale transformer retrofits with limited manpower and computing resources, achieving an efficient and cost-effective intelligent decision-making process.

[0037] 3) This invention aligns the performance score curves representing the transformer before and after the transformation by time and generates a difference curve to visually demonstrate the gains or degradations brought about by the transformation. When the difference curve tends to rise overall and fluctuates stably, it indicates that the transformation plan has made a positive contribution to efficiency, temperature rise, and noise control. If the difference is in a long-term downward trend or fluctuates violently, it indicates that there are hidden dangers in the design or process. Without additional disassembly and inspection, operation and maintenance personnel can quickly identify the reliability and potential risks of the transformation from the curve shape, and then decide to promote, optimize, or terminate subsequent work, achieving precise pre-emptive control and reducing operation and maintenance risks and economic losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The present invention will be further described below with reference to the accompanying drawings.

[0039] Figure 1 It is a flow chart of a transformer performance evaluation method based on data analysis according to the present invention. DETAILED DESCRIPTION

[0040] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making any creative efforts shall fall within the scope of protection of the present invention.

[0041] See also Figure 1 As shown, the present invention is a transformer performance evaluation method based on data analysis, comprising the following steps:

[0042] Multiple types of sensors are deployed on site to collect real-time environmental factors around the transformer, including ambient temperature, relative humidity, altitude and air pressure, etc. These raw data streams are time-synchronized and used for subsequent grouping of transformers. After grouping, the system reads the modification records in the operation and maintenance management platform, and marks the equipment with rewound windings and replaced tap changers with sealed types as one type of target transformer, and marks the equipment with only replaced oil tanks and added online monitoring interfaces as another type of target transformer, forming clear labels for the same modification methods within the group (only these three modification methods are used as examples here, and other modification methods are not listed one by one).

[0043] It is important to note that blocking the interference of the natural environment in advance allows subsequent performance comparisons to truly focus on the transformation measures themselves; uniformly marking equipment with the same transformation method after grouping can avoid repeated searches in large-scale data, help to quickly schedule computing resources and shorten the evaluation process, and at the same time lay the foundation for the subsequent selection of representative samples. Ultimately, it can help the operation and maintenance team accurately judge the actual value of different transformation plans in a complex context, reduce decision-making risks and improve management efficiency.

[0044] In a preferred embodiment of the present invention, grouping transformers includes:

[0045] Retrieve historical operation and maintenance records, collect multi-dimensional environmental curves such as temperature and humidity, salt spray concentration, and partial discharge noise, and performance indicators of the same period, slice them according to the time window, and input them into the convolutional neural network fusion attention mechanism model that has been calibrated in the laboratory. Through gradient inversion, extract the contribution of each dimension to the performance fluctuation, and automatically register the dimension whose contribution exceeds the preset threshold as the target factor; then, for these target factors, call the built-in anomaly detection module of the edge gateway to label the anomaly points, and then use sliding median filtering and wavelet soft threshold denoising to clean up burrs and high-frequency noise, and normalize the calibrated sensor signals to zero mean unit variance every hour. If data is missing for a certain period of time, the previous and next sequences of the same factor are used as input, and the trained bidirectional gated recurrent network is used to predict the reasonable value of the missing position in real time and fill in the time series library. After preprocessing, the system constructs a coordinate system with the number of target factors n plus a one-dimensional timestamp, and maps the normalized temperature and humidity A1, salt spray index A2, etc. at the same sampling time into points (A1, A2, ..., S). For all coordinate points of a single transformer within the selected observation period, a cubic B-spline fitting is used to generate a smooth curve, and the curves are aligned using the dynamic time warping algorithm to calculate the cosine similarity, and transformers with similarity higher than the threshold are classified into the same group.

[0046] It is understandable that such processing can automatically lock the environmental dimension that best explains the performance difference among massive measurement points, avoiding irrelevant features from interfering with the evaluation results; systematic cleaning, denoising and deep prediction compensation ensure the integrity and consistency of the input data, providing a reliable basis for subsequent modeling; using time coordinates to expand to n+1 dimensional space and then perform curve fitting, the discrete monitoring sequence can be converted into an object with stable geometric shape and easy to measure, and the use of high similarity threshold screening within the group further ensures that the environmental background within the group is almost the same, helping the evaluation process to eliminate the heterogeneity of external factors in the early stage, and ultimately allowing the performance difference before and after the transformation to truly focus on the technical measures themselves, such as winding rewinding, oil tank sealing or tap changer selection, providing a solid and credible objective basis for the feasibility conclusion of the transformation.

[0047] The time point t1 when the target transformer starts to be transformed and the time point t2 when the transformation is completed are obtained, a first period and a second period are set based on the time point t1 and the time point t2, and a plurality of first time points and second time points are set in the first period and the second period respectively.

[0048] In another preferred embodiment of the present invention, setting the first time point and the second time point includes:

[0049] In the asset management platform, call the API to retrieve the renovation work order for each target transformer, directly read the "Construction Start" and "Completion Acceptance" fields, and parse the dates into time points t1 and t2. The project team sets T to one month in the parameter configuration, and the system automatically generates the first cycle [t1-30d, t1] and the second cycle [t2, t2+30d] according to the length of 30 days. For example, if the renovation date is 2023-04-01 to 2023-08-15, the first cycle is 2023-03-02 to 2023-04-01, and the second cycle is 2023-08-15 to 2023-09 -14; then the system evenly slices the time axis according to the monitoring granularity: if routine monitoring is once every half day, then in the first cycle, time tags are written every half day from 2023-03-02 to 2023-04-01, forming the first time point sequence. Similarly, in the second cycle, from 2023-08-15, the same interval is calculated to 2023-09-14 to obtain the second time point sequence; the system writes the two time tags together with the transformer ID into the task queue. The subsequent scoring model uses this to pull the same amount and the same interval of operation, electrical inspection and environmental data to ensure that symmetrical samples of a full month are taken before and after the transformation.

[0050] It should be noted that the use of a one-month window can filter out sudden short-term fluctuations and avoid the load and climate mismatch caused by cross-seasonal factors, so that the comparison before and after the transformation can focus on the process changes of the equipment itself; the shortened cycle makes the data volume more streamlined, and the computing resource usage and storage pressure are reduced accordingly, which facilitates the rapid issuance of results during batch evaluation; the unified thirty-day interval and fixed sampling rhythm allow transformers of different regions and types to be on the same time scale, reducing sample voids caused by differences in operation and maintenance cycles, and improving the structural consistency of the feature matrix, thereby providing a stable and repeatable input basis for subsequent performance scoring and transformation feasibility judgment.

[0051] Performance scores of the target transformer are obtained at a first time point and a second time point, respectively recorded as a first score and a second score. The target transformer is classified based on the first score and the second score. For a single classification, a representative transformer is determined based on the first score and the second score of the target transformer in the classification. The feasibility of transformation of the target transformer is evaluated based on the first score and the second score of the representative transformer.

[0052] In another preferred embodiment of the present invention, obtaining the performance score of the target transformer includes:

[0053] After capturing the original operating data such as load current, phase voltage, oil temperature, winding hot spot temperature, gas concentration in oil, partial discharge level, vibration acceleration, etc. synchronously through the online monitoring terminal and SCADA link on the field side, the system first filters out the failed channels according to the health flag of the sensor itself, and then uses the sliding window H-ampel detection to eliminate outliers. For short-term missing segments, time series K-NN interpolation is called, and then wavelet soft threshold noise reduction is used to suppress high-frequency burrs and multi-point linear interpolation is used to align the timestamps. Finally, z-score normalization is performed according to each range; then the cleaned multi-channel signal is fed into the signal processing pipeline: the current and voltage are fast Fourier transformed to extract the fundamental amplitude The sliding mean square gradient of the temperature series is calculated to capture heat dissipation changes based on the harmonic content. The continuous wavelet transform is used to obtain the energy density distribution of the partial discharge and vibration series, and statistical indicators such as peak value, kurtosis, and spectral entropy are extracted to form an influencing parameter vector describing the winding, core, insulation, and mechanical state. Subsequently, a gradient boosting tree-gated recurrent network hybrid model is established based on historical operating conditions and maintenance records. The decision tree first partitions the static features, and then the recurrent network aggregates the dynamic features to output a comprehensive score. After offline training, the model is packaged as a service and deployed on the edge server. Whenever a new batch of influencing parameters is generated, it is pushed to the model endpoint in real time, returning the current performance score of the target transformer and writing it to the time series database for subsequent evaluation and call.

[0054] Understandably, this processing can eliminate interference caused by instrument failures and environmental noise at the data entry stage, ensuring that the input maintains structural integrity and time consistency. Systematic feature extraction compresses the complex raw signal into a small number of parameters with clear physical meanings, reducing the complexity of subsequent algorithms and providing an intuitive basis for expert review. The machine learning model integrates static operating conditions and dynamic trends, maintaining consistent evaluation criteria across different brands and load scenarios, improving the objectivity and transferability of the assessment. The final output performance score serves as a unified metric for pre- and post-modification comparisons and feasibility assessments, helping the operation and maintenance team quickly identify performance risks and process optimization opportunities during decision-making.

[0055] In a preferred embodiment of the present invention, determining the representative transformer includes:

[0056] Under the same classification, the system first assembles the first score and the second score of each target transformer into a sequence of equal length in chronological order, and calls the data visualization component to draw two broken lines at one time: A1 takes the first time point to the end, and A2 takes the second time point to the end. The horizontal axis uniformly uses the relative number of days from the transformation benchmark date to ensure the alignment of the curves; then the dynamic time warping algorithm is used to elastically align the A1 curves of the two transformers, and then the cosine similarity of the aligned sequence is calculated as the first similarity. Similarly, the operation is repeated for the A2 curve to obtain the second similarity; if the first similarity and the second similarity between the two curves exceed the second similarity threshold determined by the empirical method in the training stage, the two transformers are retained in the current classification, and all combinations are traversed in this way to output the similarity matrix; then the system performs the summation on the i-th row in the matrix, and p1 j and p2 j The selection value is accumulated pair by pair. After looping through all i, the row with the largest selection value is found. The associated transformer is written into the database and marked as the representative transformer. Its ID and selection value are recorded together for subsequent performance difference and transformation feasibility assessment.

[0057] It should be noted that a strict homogeneity screening mechanism is established using the similarity of dual-channel scoring curves to ensure highly consistent operating performance within the classification. On this basis, the device that best represents the overall curve shape is selected through a sum maximization strategy. Subsequent in-depth analysis of one device can replace repeated calculations of the entire type of equipment, thereby reducing computing power consumption and data scheduling during evaluation. At the same time, it ensures that the state changes of the representative devices can accurately map the actual transformation effects of the entire type of transformers, providing credible sample support for the subsequent conclusion of whether the difference curve rises or falls, helping operation and maintenance personnel to efficiently obtain reliable feasibility judgments in large-scale asset management scenarios.

[0058] In another preferred embodiment of the present invention, evaluating the feasibility of transformation of a target transformer includes:

[0059] Pull the complete scoring curves d1 and d2 representing the transformer at the first and second time point sequences from the database, and synchronize the two curves at t2, the day the transformation is completed. The timestamp corresponding to t2 is remapped to t=0, and the timestamp is uniformly subtracted from the remaining sampling moments, so that the two curves are aligned only on the horizontal axis without changing the vertical value. The curves after the horizontal axis synchronization are recorded as D1 and D2, and then subtracted point by point to form a difference curve D3=D2-D1. The system scans the sign change along D3, and for each interval D3<0, linear interpolation is used to determine the intersection position with the x-axis and the cumulative interval length, which is then divided by the total observation time to obtain the ratio X. At the same time, the number of intersections N of the entire D3 with the x-axis is counted. Finally, the evaluation score K is synthesized according to the weighted formula in the configuration file, and after comparison with the threshold K1, all target transformers in the classification are labeled "good feasibility" or "poor feasibility".

[0060] Aligning the time axis rather than the vertical value allows the two scoring curves to maintain their original amplitudes, accurately reflecting the absolute difference in scores before and after the renovation. The difference curve D3 directly displays the increase and decrease relationship between the two curves at the same time node. The negative area ratio X captures the persistence of the post-renovation score being lower than the pre-renovation score, and the number of intersections N characterizes the frequency of fluctuations. The combined evaluation score of the two takes into account both amplitude and stability. Using a unified threshold to convert the continuous signal analysis results into executable binary labels can quickly locate renovation batches requiring review or optimization in large-scale asset management scenarios, providing a clear and reliable basis for operation and maintenance decision-making.

[0061] A transformer performance evaluation method based on data analysis is applied to operating transformers. A remote central computer is used to collect transformer operating data through SCADA, and the data is applied to the feasibility assessment of transformer transformation.

[0062] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters and thresholds in the formulas are set by technicians in this field according to actual conditions.

[0063] The above is a detailed description of an embodiment of the present invention. However, the content is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A transformer performance evaluation method based on data analysis, characterized in that: The following steps are involved: Collect environmental factors around the transformers, group the transformers based on changes in these environmental factors, and mark transformers in the same group with the same modification method as target transformers; Obtaining a time point t1 at which the target transformer starts transformation and a time point t2 at which the transformation is completed, setting a first period and a second period based on the time point t1 and the time point t2, and setting a plurality of first time points and second time points within the first period and the second period, respectively; Obtaining performance scores of the target transformer at a first time point and a second time point, respectively recorded as a first score and a second score, classifying the target transformer based on the first score and the second score, determining a representative transformer for a single classification based on the first score and the second score of the target transformer in the classification, and evaluating the feasibility of transformation of the target transformer based on the first score and the second score of the representative transformer; Representative transformers identified include: Draw a curve A1 showing changes in the first score over time and a curve A2 showing changes in the second score over time; In the same category, the first similarity corresponding to any two target transformers is greater than the preset second similarity threshold and the second similarity corresponding to any two target transformers is greater than the preset second similarity threshold. The first similarity represents the similarity between curves A1, and the second similarity represents the similarity between curves A2; Calculate the selected value of target transformer i , p1 j and p2 j denote the first similarity and the second similarity between target transformer i and target transformer j respectively, and m denotes the total number of target transformers in the classification except target transformer i; The target transformer corresponding to the maximum selection value is taken as the representative transformer; The feasibility assessment of the target transformer modification includes: Draw a curve d1 and a curve d2 representing the change of the first score of the transformer over time, respectively, and take the time points corresponding to the starting points of the curves d1 and d2 as the origins to obtain new curves, which are recorded as curve D1 and curve D2 respectively; Get the curve D3 = D2 - D1, count the proportion X of the domain of the part below the x-axis on the curve D3 to the domain of the curve D3, and calculate the evaluation score , N represents the number of intersections of the curve D3 and the x-axis.

2. The transformer performance evaluation method based on data analysis according to claim 1, characterized in that: Grouping transformers involves: Determine target factors among environmental factors whose impact on transformer performance exceeds a preset value based on deep learning technology, preprocess the target factors, including cleaning, denoising, and normalization, and fill in missing values ​​based on predicted values ​​from a machine learning algorithm; Construct an n+1 dimensional coordinate system, where n represents the type of target factor, and generate coordinate points (A1, A2, ..., A n , S), A n represents the target factor after the nth preprocessing, and S represents the time point of collecting environmental factors; The coordinate points are fitted to obtain a fitting curve. In the same group, the similarity between the fitting curves corresponding to any two transformers is greater than a preset first similarity threshold.

3. The transformer performance evaluation method based on data analysis according to claim 1, characterized in that: Setting the first time point and the second time point includes: Set the first period [t1-T, t1] and the second period [t2, t2+T], where T represents the preset duration; Starting from the starting points of the first cycle and the second cycle respectively, a number of time points are selected within the first cycle and the second cycle at preset time intervals, and are recorded as the first time point and the second time point respectively.

4. The transformer performance evaluation method based on data analysis according to claim 1, characterized in that: Obtaining the performance score of the target transformer includes: Collecting operating data of the target transformer and performing a first processing on the operating data, the first processing including cleaning, denoising and normalization processing; Using signal processing and machine learning algorithms, extracting influencing parameters from the operating data after the first processing, wherein the influencing parameters are characteristic parameters that affect the performance state of the transformer; A transformer performance evaluation model is constructed based on the influencing parameters, the current influencing parameters are input into the transformer performance evaluation model, and the performance score of the target transformer is output.

5. The transformer performance evaluation method based on data analysis according to claim 1, characterized in that: The feasibility assessment of the target transformer modification also includes: Set the evaluation score threshold K1; If the evaluation score K≥K1, it is determined that the transformation feasibility of all target transformers in the classification is poor; otherwise, it is determined that the transformation feasibility of all target transformers in the classification is good.

6. An application of the transformer performance evaluation method based on data analysis according to any one of claims 1 to 5 in operating a transformer, characterized in that: A remote central computer is used to collect transformer operating data through SCADA and apply it to the feasibility assessment of transformer transformation.

Images