Transformer fault monitoring method and system based on artificial intelligence

Abstract

The invention relates to a transformer fault monitoring method and system based on artificial intelligence, and relates to the field of transformers, and the method comprises the steps: obtaining current data, temperature data, infrared video and load data of a transformer in a preset historical time period, and an audio oscillogram during operation; and searching a reference audio oscillogram from a preset database based on the current data, the temperature data and the load data, comparing the reference audio oscillogram with the audio oscillogram to determine an inconsistent waveform segment in the audio oscillogram, and determining the operation stability of the transformer based on the inconsistent waveform segment and the infrared video. And determining a potential fault area of the transformer according to the operation stability and the infrared video. The method has the effect of finding the accurate position of the potential fault of the transformer more timely.

Description

Technical Field

[0001] The present application relates to the field of transformers, and in particular, to a transformer fault monitoring method and system based on artificial intelligence. Background Art

[0002] A transformer is an extremely important electrical device in the power transmission and distribution system. However, due to various reasons, it may sometimes malfunction during operation. Therefore, the stable operation of the transformer is crucial. Currently, there are many pain points in transformer monitoring. At present, the transformer is usually simply monitored for temperature, and it is judged whether a fault has occurred according to the temperature change of the transformer. However, the potential fault location on the transformer cannot be detected in a timely and accurate manner at the initial stage of the fault. Therefore, how to timely detect the exact location of the potential fault during transformer fault monitoring has become a problem. Summary of the Invention

[0003] In order to more timely detect the exact location of the potential fault of the transformer, the present application provides a transformer fault monitoring method and system based on artificial intelligence.

[0004] In a first aspect, the present application provides a transformer fault monitoring method based on artificial intelligence, adopting the following technical solution: A transformer fault monitoring method based on artificial intelligence includes: Obtaining the current data, temperature data, infrared video, load data, and audio waveform diagram during operation of the transformer within a preset historical time period; Searching for a reference audio waveform diagram from a preset database based on the current data, temperature data, and load data; Comparing the reference audio waveform diagram with the audio waveform diagram to determine the inconsistent waveform segments in the audio waveform diagram; Determining the operation stability of the transformer based on the inconsistent waveform segments and the infrared video; Determining the area where the transformer may have potential faults according to the operation stability and the infrared video.

[0005] By adopting the above technical solution, current data, temperature data, infrared videos, load data, and audio waveforms during operation in a preset historical time period are obtained, which facilitates subsequent analysis of the areas where faults may occur on the transformer based on the operation of the transformer during this period. Current data, temperature data, and load data are all key factors characterizing the operating environment of the transformer. The audio waveform is the sound emitted by the transformer during operation, and the infrared video characterizes the temperature performance of each area during the operation of the transformer. That is, the audio waveform and the infrared videos of each area are the responses of the transformer in the operating environment. The area where the transformer may have a fault can be analyzed based on the audio waveform. Therefore, a reference audio waveform is found from the preset database according to the current data, etc. The reference audio waveform characterizes the sound emitted when the transformer operates normally under a situation similar to its operating environment. By comparing the reference audio waveform with the audio waveform, the inconsistent waveform segments in the audio waveform are obtained. Through comprehensive analysis of the inconsistent waveform and the infrared image, the operating stability of the transformer can be determined. The higher the operating stability, the more stable the overall operation of the transformer, and the smaller the possibility of potential fault locations. On the contrary, the possibility of fault locations is greater. Therefore, finally, based on the operating stability and the infrared image, the areas where potential faults may occur on the transformer can be analyzed, and ultimately, the precise location of potential faults in the transformer can be detected more timely.

[0006] In another possible implementation, the finding of the reference audio waveform from the preset database based on the current data, temperature data, and load data includes: Determine the current data, temperature data, and load data at the same moment; Calculate the product of the current data and the load data at the same moment, and calculate the ratio of the product to the temperature data at the corresponding moment; Calculate the first average value of all the ratios, and generate a line chart of the ratios based on all the ratios in the preset historical time period; Calculate the first similarity between the line chart and the line charts in the preset database. The preset database includes multiple audio waveforms, and each audio waveform corresponds to a line chart of ratios; Determine the target line chart with the first similarity reaching the preset similarity threshold from the line charts in the preset database; Calculate the second average value of the ratios in each target line chart, and calculate the first difference between the second average value of each target line chart and the first average value; Determine the matching value of each target line chart based on the first difference and the first similarity, and determine the audio waveform corresponding to the target line chart with the highest matching value as the reference audio waveform.

[0007] In another possible implementation manner, determining the operation stability of the transformer based on the inconsistent waveform segments and the infrared video includes: Determine the first quantity and the total duration of all the inconsistent waveform segments; Determine target reference waveform segments from the reference audio waveform diagram based on the time intervals of each inconsistent waveform segment; Determine the distance and the offset angle between the peak of each inconsistent waveform segment and the peak of the target reference waveform segment; Determine the first outlier value of the transformer based on the first quantity, the total duration, the distance, and the deviation angle; Divide the preset historical time period into multiple sub-time periods, and segment the infrared video according to the sub-time periods to obtain sub-video segments corresponding to each sub-time period; Perform grayscale transformation on each sub-video segment to obtain a corresponding sub-grayscale video segment; segment the transformer in each sub-grayscale video segment into multiple regions, and calculate the grayscale average value and the grayscale variance of each region in each sub-grayscale video segment; Determine the second outlier value of the transformer based on the grayscale average value and the grayscale variance of each region; Determine the operation stability of the transformer based on the first outlier value and the second outlier value.

[0008] In another possible implementation manner, determining the first outlier value of the transformer based on the first quantity, the total duration, the distance, and the deviation angle includes: Calculate the deviation score of each inconsistent waveform segment based on the distance and the offset angle; Determine candidate waveform segments whose deviation scores reach a preset score threshold, and calculate the average value of the deviation scores of all the candidate waveform segments; Calculate the first outlier value based on the first quantity, the total duration, the deviation score, and their respective corresponding coefficients.

[0009] In another possible implementation manner, determining the area where the transformer may have potential faults based on the operation stability and the infrared video includes: Determine the maximum grayscale value of each region in each sub-grayscale video segment, and calculate the second difference between the maximum grayscale value and the grayscale average value of the region; Determine the second quantity of each region where the second difference reaches a preset difference threshold in all the sub-grayscale video segments; Calculate the ratio of the second quantity to the quantity of all the sub-grayscale video segments; Calculate the third difference between the grayscale average value of each sub-grayscale video segment of each region and the preset average value threshold of each region; Calculate the average value of the third difference based on all the third differences in each region; Determine the fault value of each region based on the proportion and the average value of the third difference; Multiply the operating stability by the fault value of each region to obtain the probability of a fault occurring in each region; Determine the region where the probability reaches the preset probability threshold as the region where the transformer may have potential faults.

[0010] In another possible implementation, the determining the second outlier of the transformer based on the average gray value and gray variance of each region in each region includes: Determine the target variance of all the average gray values based on the average gray value of all the sub-gray video segments in each region; Determine the average value of all the gray variances based on the gray variance of all the sub-gray video segments in each region; Calculate the second outlier based on the target variance, the average value of all the gray variances, and their respective corresponding coefficients.

[0011] In another possible implementation, the method further includes: Send the region where potential faults may occur to the terminal device of the staff.

[0012] In a second aspect, the present application provides an artificial intelligence-based transformer fault monitoring system, adopting the following technical solutions: An artificial intelligence-based transformer fault monitoring system includes: A data acquisition module, configured to acquire the current data, temperature data, infrared video, load data, and audio waveform diagram during operation of the transformer within a preset historical time period; A search module, configured to search for a reference audio waveform diagram from a preset database based on the current data, temperature data, and load data; A comparison module, configured to compare the reference audio waveform diagram with the audio waveform diagram to determine the inconsistent waveform segments in the audio waveform diagram; A stability determination module, configured to determine the operating stability of the transformer based on the inconsistent waveform segments and the infrared video; A region determination module, configured to determine the region where the transformer may have potential faults according to the operating stability and the infrared video.

[0013] By adopting the above technical solution, the data acquisition module acquires current data, temperature data, infrared videos, load data, and audio waveforms during operation in a preset historical time period, which facilitates subsequent analysis of possible fault areas on the transformer based on the operation conditions of the transformer during this period. The current data, temperature data, and load data are all key factors characterizing the operation environment of the transformer. The audio waveform is the sound emitted by the transformer during operation, and the infrared video characterizes the temperature performance of each area during the operation of the transformer. That is, the audio waveform and the infrared videos of each area are the responses of the transformer in the operation environment. According to the audio waveform, the possible fault areas of the transformer can be analyzed. Therefore, the search module finds a reference audio waveform from a preset database based on the current data, etc. The reference audio waveform characterizes the sound emitted during the normal operation of the transformer under conditions similar to its operation environment. The comparison module compares the reference audio waveform with the audio waveform to obtain inconsistent waveform segments in the audio waveform. The stability determination module comprehensively analyzes the inconsistent waveform and the infrared image to determine the operation stability of the transformer. The higher the operation stability, the more stable the overall operation of the transformer, and the lower the possibility of potential fault locations. On the contrary, the possibility of fault locations is high. Therefore, finally, the area determination module can analyze the possible potential fault areas on the transformer based on the operation stability and the infrared image, and finally achieve more timely discovery of the precise locations of potential faults on the transformer.

[0014] In another possible implementation manner, when the search module searches for a reference audio waveform from a preset database based on the current data, temperature data, and load data, it is specifically used for: Determine the current data, temperature data, and load data at the same time; Calculate the product of the current data and the load data at the same time, and calculate the ratio of the product to the temperature data at the corresponding time; Calculate the first average value of all the ratios, and generate a line chart of the ratios based on all the ratios in the preset historical time period; Calculate the first similarity between the line chart and the line charts in the preset database. The preset database includes multiple audio waveforms, and each audio waveform corresponds to a line chart of ratios; Determine the target line chart with the first similarity reaching the preset similarity threshold from the line charts in the preset database; Calculate the second average value of the ratios in each target line chart, and calculate the first difference between the second average value of each target line chart and the first average value; Determine the matching value of each target line chart based on the first difference and the first similarity, and determine the audio waveform corresponding to the target line chart with the highest matching value as the reference audio waveform.

[0015] In another possible implementation, when determining the operation stability of the transformer based on the inconsistent waveform segments and the infrared video, the stability determination module is specifically configured to: Determine the first quantity and the total duration of all the inconsistent waveform segments; Determine target reference waveform segments from the reference audio waveform diagram based on the time intervals of each inconsistent waveform segment; Determine the distance and the offset angle between the wave crest of each inconsistent waveform segment and the wave crest of the target reference waveform segment; Determine a first outlier value of the transformer based on the first quantity, the total duration, the distance, and the deviation angle; Divide the preset historical time period into multiple sub-time periods, and segment the infrared video according to the sub-time periods to obtain sub-video segments corresponding to each sub-time period; Perform grayscale transformation on each sub-video segment to obtain a corresponding sub-grayscale video segment, segment the transformer in each sub-grayscale video segment into multiple regions, and calculate the grayscale average value and the grayscale variance of each region in each sub-grayscale video segment; Determine a second outlier value of the transformer based on the grayscale average value and the grayscale variance of each region; Determine the operation stability of the transformer based on the first outlier value and the second outlier value.

[0016] In another possible implementation, when determining the first outlier value of the transformer based on the first quantity, the total duration, the distance, and the deviation angle, the stability determination module is specifically configured to: Calculate the deviation score of each inconsistent waveform segment based on the distance and the offset angle; Determine candidate waveform segments whose deviation scores reach a preset score threshold, and calculate the average value of the deviation scores of all the candidate waveform segments; Calculate the first outlier value based on the first quantity, the total duration, the deviation score, and their respective coefficients.

[0017] In another possible implementation, when determining the area where potential faults may occur in the transformer based on the operation stability and the infrared video, the area determination module is specifically configured to: Determine the maximum grayscale value of each region in each sub-grayscale video segment, and calculate a second difference value between the maximum grayscale value and the grayscale average value of the region; Determine a second quantity of each region where the second difference value reaches a preset difference threshold in all the sub-grayscale video segments; Calculate the ratio of the second quantity to the quantity of all the sub-grayscale video segments; Calculate the third difference between the grayscale average value of each sub-grayscale video segment in each area and the preset average value threshold of each area; Calculate the average value of the third differences based on all the third differences of each area; Determine the fault value of each area based on the proportion and the average value of the third differences; Multiply the operation stability by the fault value of each area to obtain the probability of a fault occurring in each area; Determine the area where the probability reaches the preset probability threshold as the area where the transformer may have potential faults.

[0018] In another possible implementation, when the stability determination module determines the second outlier of the transformer based on the grayscale average value and grayscale variance of each area, it is specifically configured to: Determine the target variance of all the grayscale average values based on the grayscale average value of all the sub-grayscale video segments in each area; Determine the average value of all the grayscale variances based on the grayscale variance of all the sub-grayscale video segments in each area; Calculate the second outlier based on the target variance, the average value of all the grayscale variances, and their respective corresponding coefficients.

[0019] In another possible implementation, the transformer fault monitoring system based on artificial intelligence further includes: A sending module, configured to send the area where potential faults may occur to the terminal device of the staff.

[0020] In a third aspect, the present application provides an electronic device, adopting the following technical solution: An electronic device, the electronic device includes: At least one processor; A memory; At least one application program, where at least one application program is stored in the memory and is configured to be executed by at least one processor, and at least one is configured to: execute a transformer fault monitoring method based on artificial intelligence as shown in any possible implementation manner of the first aspect.

[0021] In a fourth aspect, the present application provides a computer-readable storage medium, adopting the following technical solution: A computer-readable storage medium, when the computer program is executed in a computer, causes the computer to execute the transformer fault monitoring method based on artificial intelligence as described in any item of the first aspect.

[0022] In summary, the present application includes at least one of the following beneficial technical effects: Obtain the current data, temperature data, infrared video, load data, and audio waveform diagram during operation in a preset historical time period, which is convenient for subsequently analyzing the areas where faults may occur on the transformer based on the operation conditions of the transformer during this period. The current data, temperature data, and load data are all key factors characterizing the operating environment of the transformer. The audio waveform diagram is the sound emitted by the transformer during operation, and the infrared video characterizes the temperature performance of each area during the operation of the transformer. That is, the audio waveform diagram and the infrared video of each area are the responses of the transformer under the operating environment. According to the audio waveform diagram, the areas where the transformer may have faults can be analyzed. Therefore, based on the current data, etc., find the reference audio waveform diagram from the preset database. The reference audio waveform diagram characterizes the sound emitted when the transformer operates normally under a situation similar to its operating environment. Compare the reference audio waveform diagram with the audio waveform diagram to obtain the inconsistent waveform segments in the audio waveform diagram. By comprehensively analyzing the inconsistent waveform diagram and the infrared image, the operating stability of the transformer can be determined. The higher the operating stability, the more stable the overall operation of the transformer, and the smaller the possibility of potential fault locations. On the contrary, the possibility of fault locations is greater. Therefore, finally, based on the operating stability and the infrared image, the areas where potential faults may occur on the transformer can be analyzed, and ultimately the precise location of potential faults in the transformer can be discovered more timely. Description of the Drawings

[0023] Figure 1 It is a schematic flowchart of a transformer fault monitoring method based on artificial intelligence according to an embodiment of the present application.

[0024] Figure 2 It is a schematic structural diagram of a transformer fault monitoring system based on artificial intelligence according to an embodiment of the present application.

[0025] Figure 3 It is a schematic structural diagram of an electronic device according to an embodiment of the present application. Detailed Description of the Embodiment

[0026] The following further describes the present application in detail with reference to the accompanying drawings.

[0027] After reading this specification, those skilled in the art can make modifications to this embodiment without creative contributions as needed, but as long as they are within the scope of the claims of the present application, they are protected by the patent law.

[0028] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the following will clearly and completely describe the technical solutions in the embodiments of this application with reference to the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application, rather than all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of this application without creative efforts shall fall within the scope of protection of this application.

[0029] In addition, the term "and / or" in this document is merely an association relationship describing associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, unless otherwise specified.

[0030] The following further describes the embodiments of this application in detail with reference to the drawings of the specification.

[0031] The embodiments of this application provide a transformer fault monitoring method based on artificial intelligence, which is executed by an electronic device. The electronic device can be a server or a terminal device. Among them, the server can be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal device can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc., but is not limited thereto. The terminal device and the server can be directly or indirectly connected through wired or wireless communication methods, which are not limited in the embodiments of this application. As Figure 1 shown, the method includes steps S101, S102, S103, S104, and S105, where S101, obtain the current data, temperature data, infrared video, load data, and audio waveform diagram during operation of the transformer within a preset historical time period.

[0032] For the embodiments of the present application, the preset historical time period can be the past half hour, the past hour, etc. of the current time. Current sensors, temperature sensors, load sensors, pickups or microphones are provided on the transformer, so as to be able to collect the current data, temperature data, load data and the sound emitted during operation of the transformer. In other embodiments, the load data can also be obtained by calculating data such as current and voltage. The temperature sensor can be arranged inside the transformer to collect the temperature inside the transformer, and then use this temperature to characterize the operating environment of the transformer. An infrared camera is also arranged in the area near the transformer to collect the infrared video of the whole transformer, and the infrared video records the specific temperatures of various parts of the whole device during the operation of the transformer. The above-mentioned sensors and other acquisition devices are connected to the electronic device through wires or wirelessly, so that the electronic device can obtain the above data. The above-mentioned sensors and other acquisition devices can also be connected to the cloud server wirelessly to upload the above data to the cloud server, and then the electronic device obtains it from the cloud server.

[0033] The current data, temperature data and load data represent the operating environment of the transformer, and the audio waveform diagram and infrared video represent the specific reactions and performance during the operation of the transformer.

[0034] S102, search for a reference audio waveform diagram from a preset database based on the current data, temperature data and load data.

[0035] For the embodiments of the present application, a plurality of audio waveform diagrams are stored in the preset database, and each audio waveform diagram is the sound emitted when the transformer operates normally under different operating environments. Since the current data, temperature data and load data represent the operating environment of the transformer, the electronic device can find a reference audio waveform diagram similar to the operating environment of the transformer during the preset historical time period from the preset database according to the current data, temperature data and load data. Comparing the reference audio waveform diagram with the audio waveform diagram within the preset time period is convenient for analyzing the fault condition of the transformer.

[0036] S103, compare the reference audio waveform diagram with the audio waveform diagram to determine the inconsistent waveform segments in the audio waveform diagram.

[0037] For the embodiments of the present application, the electronic device overlaps and compares the reference audio waveform diagram with the audio waveform diagram within the preset time period, so as to be able to eliminate the overlapping part and obtain inconsistent waveform segments. Analyzing the inconsistent waveform segments can obtain the abnormal condition and operating stability condition of the transformer.

[0038] S104, determine the operating stability of the transformer based on the inconsistent waveform segments and the infrared video.

[0039] For the embodiments of the present application, the inconsistent waveform diagram represents the difference between the sound emitted by the transformer during operation and that during normal operation, thus affecting the operation stability of the transformer. The infrared video records the temperature performance at different positions in different regions of the transformer, and the positions with abnormal temperature affect the stable operation of the transformer. Therefore, the electronic device can accurately obtain the operation stability of the transformer by comprehensively analyzing the inconsistent waveform segment and the infrared video.

[0040] S105. Determine the area where the transformer may have potential faults according to the operation stability and the infrared video.

[0041] For the embodiments of the present application, the level of operation stability affects the probability of faults occurring on the transformer. Combining with the infrared video can determine the change in temperature performance at different positions in different regions of the transformer. Combining the overall operation stability of the transformer and the infrared video can accurately determine the area where the transformer may have potential faults, so as to more timely discover the precise position where potential faults may occur on the transformer.

[0042] In a possible implementation manner of the embodiments of the present application, in step S102, the reference audio waveform diagram is searched from the preset database based on the current data, temperature data, and load data, which specifically includes step S1021 (not shown in the figure), step S1022 (not shown in the figure), step S1023 (not shown in the figure), step S1024 (not shown in the figure), step S1025 (not shown in the figure), step S1026 (not shown in the figure), and step S1027 (not shown in the figure). Among them, S1021. Determine the current data, temperature data, and load data at the same moment.

[0043] For the embodiments of the present application, the current data, temperature data, and load data in the preset time period can all be represented on the coordinate axis, thus forming their respective line charts. The respective line charts are placed together for comparison to determine the current data, temperature data, and load data at each same moment.

[0044] S1022. Calculate the product of the current data and the load data at the same moment, and calculate the ratio of the product to the temperature data at the corresponding moment.

[0045] For the embodiments of the present application, the current data, temperature data, and load data at the same time characterize the operating environment of the transformer at that moment. The current data and load data are the impacts of the power grid distribution on the operating environment, and the temperature data is the impact of the transformer itself on the operating environment. Therefore, the electronic device calculates the product of the current data and the load data at the same time to obtain a value, and uses this value to characterize the comprehensive impact of the current and the load on the transformer. Then, the product is divided by the temperature data at the corresponding moment to obtain a ratio, and this ratio characterizes the operating environment at that moment. Different operating environments correspond to different ratios. Therefore, this ratio is the characteristic value of the operating environment at each moment. By using the ratio at each moment, it is convenient to subsequently find a reference audio waveform diagram that is closer to the operating environment in the preset historical period.

[0046] S1023, calculate the first average value of all the ratios, and generate a line graph of the ratios based on all the ratios in the preset historical period.

[0047] For the embodiments of the present application, the electronic device calculates the first average value of all the ratios through the average value calculation formula. The first average value characterizes the level of the operating environment, which is convenient for subsequently finding a reference audio waveform diagram with the same or similar level. The electronic device maps the ratios and the corresponding moments to the preset plane rectangular coordinate system, and then connects them in sequence to obtain a line graph of the ratios.

[0048] S1024, calculate the first similarity between the line graph and the line graph in the preset database.

[0049] Among them, the preset database includes multiple audio waveform diagrams, and each audio waveform diagram corresponds to a line graph of the ratios.

[0050] For the embodiments of the present application, the electronic device can calculate the first similarity between the line graph and the line graph in the preset database by calculating the cosine similarity, Euclidean distance, Pearson correlation coefficient, etc. The higher the first similarity, the closer the change trend of the operating environment over time.

[0051] S1025, determine the target line graph in the line graph in the preset database whose first similarity reaches the preset similarity threshold.

[0052] For the embodiments of the present application, the preset similarity threshold represents the demarcation point of whether the first similarity is high. For example, the preset similarity threshold can be 95%. Reaching the preset similarity threshold indicates that the change trend of the operating environment over time is relatively close. Therefore, the electronic device uses 95% for screening to obtain the target line graph that reaches the preset similarity threshold, and further screens out the most suitable reference audio waveform diagram from the target line graph.

[0053] S1026. Calculate the second average value of the ratios in each target line chart, and calculate the first difference between the second average value and the first average value of each target line chart.

[0054] For the embodiments of the present application, the second average value is calculated according to the average value calculation formula for the ratios at each moment in each target line chart, and then the first difference is obtained by subtracting the first average value from the second average value. The electronic device can take the absolute value of the first difference. The smaller the absolute value, the closer the level of the operating environment, and the higher the first similarity, the more suitable it is to be regarded as a similar operating environment.

[0055] S1027. Determine the matching value of each target line chart based on the first difference and the first similarity, and determine the audio waveform chart corresponding to the target line chart with the highest matching value as the reference audio waveform chart.

[0056] For the embodiments of the present application, both the first difference and the first similarity are key factors affecting the matching degree between the operating environment of the target line chart and the operating environment of the transformer in the preset historical time period. Therefore, different coefficients are set for each of the first difference and the first similarity, and then the electronic device performs weighted calculation according to their respective corresponding coefficients to obtain the matching value. Since the first difference is inversely proportional to the matching degree, the electronic device can take the reciprocal of the absolute value of the first difference for weighted calculation, so that the matching value is more in line with the actual logic. The higher the matching value, the closer the operating environment, and the more suitable the corresponding audio waveform chart is as the reference audio waveform chart. By determining the characteristic values about the operating environment through the relevant data in the preset historical time period, and further determining the line chart, the reference audio waveform chart found through the line chart is more accurate.

[0057] In a possible implementation manner of the embodiments of the present application, in step S104, the operation stability of the transformer is determined based on the inconsistent waveform segments and the infrared video, which specifically includes step S1041 (not shown in the figure), step S1042 (not shown in the figure), step S1043 (not shown in the figure), step S1044 (not shown in the figure), step S1045 (not shown in the figure), step S1046 (not shown in the figure), step S1047 (not shown in the figure), and step S1048 (not shown in the figure), where S1041. Determine the first quantity and the total duration of all the inconsistent waveform segments.

[0058] For the embodiments of the present application, the more the first quantity of the inconsistent waveform diagrams, the more unstable the operation of the transformer. The electronic device sums up the durations occupied by all the inconsistent waveform segments to obtain the total duration. The longer the total duration, the longer the time when the operation of the transformer is inconsistent with the sound emitted during normal operation, and it also indicates that the operation is more unstable.

[0059] S1042, determine a target reference waveform segment from the reference audio waveform diagram based on the time interval of each inconsistent waveform segment.

[0060] For the embodiments of the present application, each inconsistent waveform segment corresponds to a start time and an end time, and the time interval between the start time and the end time is the time interval of the inconsistent waveform segment. The electronic device can segment the target reference waveform segment corresponding to each inconsistent waveform segment from the reference audio waveform diagram according to the time interval.

[0061] S1043, determine the distance and the offset angle between the peak of each inconsistent waveform segment and the peak of the target reference waveform segment.

[0062] For the embodiments of the present application. By comparing the inconsistent waveform segment and the target reference waveform segment, the difference between the sound emitted when the transformer is operating suspiciously and the sound emitted when it is operating normally can be compared. The electronic device can map the inconsistent waveform segment and the target reference waveform segment to a coordinate system, and then use the distance formula between two points to calculate the distance between the peak of the inconsistent waveform segment and the peak of the target reference waveform segment. The greater the distance, the greater the difference between the sound emitted during operation and the sound emitted during normal operation. The electronic device determines the coordinate points of the two peaks in the coordinate system, and then calculates the offset angle between the two peaks. The greater the offset angle, the greater the difference between the sound emitted when the transformer is operating and the sound emitted when it is operating normally. Therefore, both the distance and the offset angle can illustrate the operating stability of the transformer. In other embodiments, the electronic device can also calculate the similarity between the inconsistent waveform segment and the target reference waveform segment, and the magnitude of this similarity also affects the operating stability of the transformer.

[0063] S1044, determine the first abnormal value of the transformer based on the first quantity, the total duration, the distance, and the deviation angle.

[0064] For the embodiments of the present application, in summary, the first quantity, the total duration, the distance, and the deviation angle are all key factors affecting the operating stability and abnormal conditions of the transformer, and these factors all characterize the operating stability and abnormal conditions from the audio perspective. Therefore, the electronic device comprehensively determines the more accurate first abnormal value of the transformer in terms of audio based on the above four factors.

[0065] S1045, divide a preset historical time period into multiple sub-time periods, and segment the infrared video according to the sub-time periods to obtain sub-video segments corresponding to each sub-time period.

[0066] For the embodiments of the present application, the electronic device can evenly divide the preset historical time period into multiple sub-time periods. Then, segment the infrared video according to each sub-time period to obtain the sub-video segments of each sub-time period.

[0067] S1046. Perform grayscale transformation on each sub-video segment to obtain the corresponding sub-grayscale video segment. Divide the transformer in each sub-grayscale video segment into multiple regions, and calculate the grayscale average value and grayscale variance of each region in each sub-grayscale video segment.

[0068] For the embodiments of the present application, the electronic device can perform denoising processing on each sub-video segment to obtain the denoised sub-video segment, and then perform grayscale transformation on the denoised sub-video segment to obtain the corresponding sub-grayscale video segment. The magnitude of the grayscale value represents the temperature. A model of the transformer can be stored in the electronic device. The staff can divide it into multiple regions on the model. The regions can be equally divided or divided according to requirements, functional modules, etc. Then the electronic device processes the grayscale values in each frame of each sub-grayscale video segment, thereby calculating the grayscale average value and grayscale variance of each region in each frame of the picture. Then, in combination with the grayscale average value and grayscale variance of all the frame pictures of the entire sub-video segment, the grayscale average value and grayscale variance of each region in each sub-grayscale video segment are calculated. The grayscale average value represents the average temperature level of each region during the time period where the sub-grayscale video segment is located, and the grayscale variance represents the difference in temperature distribution of each region within the sub-grayscale video segment.

[0069] S1047. Determine the second outlier of the transformer based on the grayscale average value and grayscale variance of each region.

[0070] For the embodiments of the present application, the grayscale average value represents the average temperature level, that is, it represents the heat generation situation of each region. If the average temperature of a certain region is too high, it indicates that the possibility of an abnormality in that region on the transformer is relatively large, which in turn affects the operating stability of the transformer. The grayscale variance represents the difference in temperature distribution within each region. If the difference in temperature distribution is too large, it indicates that the possibility of an abnormality in that region on the transformer is relatively large, which in turn affects the operating stability of the transformer. That is, the grayscale average value and grayscale variance are the influencing factors of the overall operating stability of the transformer in terms of temperature performance of each region. Therefore, the electronic device comprehensively analyzes the grayscale average value and grayscale variance to obtain an accurate second outlier.

[0071] S1048. Determine the operating stability of the transformer based on the first outlier and the second outlier.

[0072] For the embodiments of the present application, after the electronic device determines the first outlier and the second outlier, if both the first outlier and the second outlier are directly proportional to the running stability, that is, the higher the running stability is when these two outliers are larger, the sum of the first outlier and the second outlier can be directly calculated to obtain the total value, and this total value can represent the running stability of the transformer. If both the first outlier and the second outlier are inversely proportional to the running stability, that is, the lower the running stability is when these two outliers are larger, the reciprocals of the first outlier and the second outlier can be summed to obtain the running stability. The running stability of the transformer determined comprehensively from both the audio performance and the temperature performance is more accurate.

[0073] In a possible implementation manner of the embodiments of the present application, in step S1044, the first outlier of the transformer is determined based on the first quantity, the total duration, the distance, and the deviation angle, which specifically includes step Sa (not shown in the figure), step Sb (not shown in the figure), and step Sc (not shown in the figure), where Sa, calculate the deviation score of each inconsistent waveform segment based on the distance and the offset angle.

[0074] For the embodiments of the present application, both the distance and the offset angle are key factors affecting the degree of difference between the inconsistent waveform and the waveform during normal operation within the corresponding time period, and the influencing degrees are different. Therefore, the staff can set different coefficients for the distance and the offset angle and store them in the electronic device, and then the electronic device can call their respective corresponding coefficients to perform weighted calculation on the distance and the offset angle to obtain the deviation score of each inconsistent waveform.

[0075] Sb, determine the candidate waveform segments whose deviation scores reach the preset score threshold, and calculate the average value of the deviation scores of all the candidate waveform segments.

[0076] For the embodiments of the present application, the preset score threshold is the demarcation point for whether the deviation score is too large. Reaching the preset score threshold indicates that it has a greater impact on the running stability of the transformer, and not reaching the preset score threshold indicates that the inconsistent waveform segment has a relatively small impact on the running stability of the transformer. Therefore, according to the preset score threshold, a part of the inconsistent waveform segments with relatively small impact on the running stability is filtered out to obtain the candidate waveform segments that are sufficiently influential on the running stability of the transformer. Then the electronic device calculates the average value of the deviation scores of all the candidate waveform segments using the average value calculation formula. The larger the average value of the deviation scores is, the worse the running stability of the transformer is, and the greater the probability of abnormality is. Analyzing the candidate waveform segments reduces the amount of analysis compared to analyzing all the inconsistent waveform segments.

[0077] Sc, calculate the first outlier based on the first quantity, the total duration, the deviation score, and their respective corresponding coefficients.

[0078] For the embodiments of the present application, in summary, the first quantity, the total duration, and the deviation score are all key factors affecting the stable operation of the transformer in terms of audio, and the degrees of influence are different. Therefore, the staff can set their respective corresponding coefficients for the above three factors and store them in the electronic device, and the electronic device calls their respective corresponding coefficients to perform weighted calculations on the first quantity, the total duration, and the deviation score to obtain the first outlier. Calculating the first outlier through features such as the first quantity is more accurate. In other embodiments, the first quantity can also be replaced by the number of candidate waveform segments, and the total duration can also be replaced by the ratio of the total duration to the preset historical time period.

[0079] In a possible implementation manner of the embodiments of the present application, in step S105, according to the operation stability and the infrared video, the area where the transformer may have potential faults is determined, specifically including step S1051 (not shown in the figure), step S1052 (not shown in the figure), step S1053 (not shown in the figure), step S1054 (not shown in the figure), step S1055 (not shown in the figure), step S1056 (not shown in the figure), step S1057 (not shown in the figure), and step S1058 (not shown in the figure), where S1051, determine the maximum gray value of each area in each sub-gray video segment, and calculate the second difference between the maximum gray value and the average gray value of the area.

[0080] For the embodiments of the present application, the temperature is the highest at the maximum gray value. The higher the maximum gray value of a certain area, the greater the probability that the area has abnormalities and potential faults. The electronic device subtracts the average gray value of the area from the maximum gray value to obtain the second difference. The greater the second difference, the greater the temperature difference from the general temperature of the area, and the greater the probability of belonging to an abnormality or potential fault.

[0081] S1052, determine the second quantity of each area where the second difference reaches the preset difference threshold in all sub-gray video segments.

[0082] For the embodiments of the present application, the preset difference threshold is used as the demarcation point for whether the second difference is too large. If the second difference reaches the preset difference threshold, it indicates that the maximum gray value is more likely to be an abnormality or potential fault. Therefore, the electronic device determines the sub-gray video segments in which the second difference reaches the preset difference threshold in all sub-gray video segments of each area, and determines the number of these sub-gray video segments, that is, the second quantity. The more the number of these sub-gray video segments, the longer the time when there are positions with too high temperature in the area, and thus the higher the probability that the area has abnormalities and potential faults.

[0083] S1053, calculate the ratio of the second quantity to the number of all sub-gray video segments.

[0084] For the embodiments of the present application, the electronic device divides the second quantity by the quantity of all sub-gray video segments to obtain a ratio. The larger the ratio, the longer and more frequent the time of the position with too high temperature in this area, and the greater the probability of abnormal and potential faults. Using the ratio to characterize the probability of abnormal and potential faults in the area is more accurate.

[0085] S1054. Calculate the third difference between the gray average value of each sub-gray video segment in each area and the preset average value threshold of each area.

[0086] For the embodiments of the present application, the preset average value threshold is used as the temperature threshold for the overall temperature performance of each area during the normal operation of the transformer. The electronic device subtracts the corresponding preset average value threshold from the gray average value of each sub-gray video segment in each area to obtain the third difference for each sub-gray video segment. The larger the third difference, the greater the gap between the temperature performance of the area and the temperature performance during the normal operation of the transformer, and the greater the possibility of abnormal and potential faults in the area.

[0087] S1055. Calculate the average value of the third differences based on all the third differences of each area.

[0088] For the embodiments of the present application, after the electronic device determines all the third differences of each area, it uses the average value calculation formula to calculate the average value of the third differences of each area, and uses the average value of the third differences to characterize the overall gap between the temperature performance of each area during the preset historical time period and the temperature performance of the area during the normal operation of the transformer. The larger the average value of the third differences, the greater the possibility of abnormal and potential faults.

[0089] S1056. Determine the fault value of each area based on the ratio and the average value of the third differences.

[0090] For the embodiments of the present application, both the ratio and the average value of the third differences are key factors affecting the occurrence of abnormal and potential faults in each area. Therefore, the electronic device can comprehensively determine the fault value of each area according to the ratio and the average value of the third differences, that is, it is more accurate to determine the fault value according to the infrared video (temperature performance) of each area during the preset historical time period. Specifically, the staff can set the corresponding coefficients for the ratio and the average value of the third differences and store them in the electronic device, and then the electronic device calls the corresponding coefficients for weighted calculation to obtain the fault value of each area.

[0091] S1057. Multiply the operation stability by the fault value of each area to obtain the probability of a fault occurring in each area.

[0092] For the embodiments of the present application, the possibility of anomalies and potential failures occurring in each region is related not only to the temperature performance of the region itself, but also to the overall operating stability of the transformer. The more unstable the overall operating condition of the transformer is, the more accurate the fault value determined by superimposing the temperature performance of each region will be, thereby making the probability of potential failures in each region more accurate. Therefore, the electronic device can calculate the probability of potential failures in each region by multiplying the operating stability of the transformer by the fault value of each region. Specifically, the electronic device can normalize the product of the operating stability and the fault value, and map the product of each region to between 0 and 1 to obtain the probability of each region.

[0093] S1058, determine the region where the probability reaches the preset probability threshold as the region where the transformer may have potential failures.

[0094] For the embodiments of the present application, the preset probability threshold is used as the demarcation point for potential failures. Assuming the preset probability threshold is 70%, the electronic device can compare the probability of potential failures in each region with 70%. The region that reaches the preset probability threshold is the region where the transformer may have potential failures. By superimposing the temperature performance of each region and the overall operating stability of the transformer, the region with potential failures can be determined more accurately.

[0095] Furthermore, since the components included in different regions of the transformer are different, and the corresponding functions and roles are also different, the electronic device can determine the priority of each region according to the number of components in each region and the importance of the functions achieved. After the electronic device determines the region with potential failures, it comprehensively determines the danger level of the transformer according to the priority of the region with potential failures and the probability of failure in this region. When multiple regions with potential failures are determined, the danger level of the transformer is determined according to the priorities, failure probabilities, and the total number of regions with potential failures of the multiple regions with potential failures. Then, according to the power consumption change diagram of the covered management area of each transformer, the upcoming nearest power consumption trough is determined, and the nearest power consumption trough time period is determined as the maintenance time of the transformer. Or, according to the size and average power consumption of the covered management area of each transformer, the importance level of the transformer is determined. When there are multiple regions with potential failures in multiple transformers, the overhaul priority order of each transformer is determined according to the importance level and danger level of each transformer.

[0096] In a possible implementation manner of the embodiments of the present application, in step S1047, determining the second anomaly value of the transformer based on the average gray value and gray variance of each region specifically includes step one, step two, and step three, where Step one, determine the target variance of all average gray values based on the average gray value of each region for all sub-gray video segments.

[0097] For the embodiments of the present application, the electronic device calculates the target variance of each region with respect to the average grayscale value of all regions through a variance calculation formula. The target variance characterizes the fluctuation of the temperature performance of each region compared to the temperature performance of the region during normal operation. The larger the target variance, the greater the difference and fluctuation in the temperature performance of the region compared to the temperature performance during normal operation, and the worse the operating stability of the transformer.

[0098] Step 2: Determine the average value of all grayscale variances based on the grayscale variances of each region with respect to all sub-grayscale video segments.

[0099] For the embodiments of the present application, the grayscale variance characterizes the temperature difference between different positions within a region during a sub-grayscale video segment. The smaller the temperature difference between different positions within the region, the more evenly distributed the heat generation within the region, and thus the higher the operating stability of the transformer. Therefore, the electronic device calculates the average value of all grayscale variances by averaging all the grayscale variances within a preset historical time period for each region. The larger the average value of the grayscale variance of a certain region, the greater the temperature change difference between different positions within the region during the preset historical time period, the more uneven the heat generation, and the worse the operating stability of the transformer.

[0100] Step 3: Calculate the second outlier based on the target variance, the average value of all grayscale variances, and their respective coefficients.

[0101] For the embodiments of the present application, in summary, both the target variance and the average value of all grayscale variances are key factors affecting the operating stability of the transformer in terms of temperature performance. The electronic device can sum up the target variances of all regions to obtain the total target variance, and sum up the average values of the grayscale variances of all regions to obtain the total average value. The staff pre-sets the coefficients of these two totals and stores them in the electronic device. The electronic device calls their respective coefficients for weighted calculation to obtain the second outlier. Calculating the second outlier of the transformer in terms of temperature performance through data such as the target variance is more accurate.

[0102] It should be noted that the coefficients and thresholds in this embodiment can be adaptively modified according to actual situations and requirements.

[0103] In a possible implementation manner of the embodiments of the present application, after step S105, there is also step S106 (not shown in the figure), where S106: Send the regions where potential failures may occur to the terminal device of the staff.

[0104] For the embodiments of the present application, the terminal device of the staff can be a telephone, a personal computer, or other devices. The electronic device is wirelessly connected to the terminal device of the staff, so that the electronic device can determine the area where the transformer may have potential faults and send it to the terminal device of the staff, enabling the staff to timely know the operation status of the transformer.

[0105] The above embodiments introduce a transformer fault monitoring method based on artificial intelligence from the perspective of the method flow. The following embodiments introduce a transformer fault monitoring system 20 based on artificial intelligence from the perspective of virtual modules or virtual units. For details, see the following embodiments.

[0106] The embodiments of the present application provide a transformer fault monitoring system 20 based on artificial intelligence, as Figure 2 shown. A transformer fault monitoring system 20 based on artificial intelligence may specifically include: A data acquisition module 201, configured to acquire the current data, temperature data, infrared video, load data, and audio waveform diagram during operation of the transformer within a preset historical time period; A search module 202, configured to search for a reference audio waveform diagram from a preset database based on the current data, temperature data, and load data; A comparison module 203, configured to compare the reference audio waveform diagram with the audio waveform diagram to determine the inconsistent waveform segments in the audio waveform diagram; A stability determination module 204, configured to determine the operation stability of the transformer based on the inconsistent waveform segments and the infrared video; An area determination module 205, configured to determine the area where the transformer may have potential faults according to the operation stability and the infrared video.

[0107] An embodiment of the present application discloses a transformer fault monitoring system 20 based on artificial intelligence. Among them, the data acquisition module 201 acquires current data, temperature data, infrared videos, load data, and audio waveform diagrams during operation in a preset historical time period, which is convenient for subsequently analyzing the areas where faults may occur on the transformer according to the operating conditions of the transformer during this period. Current data, temperature data, and load data are all key factors characterizing the operating environment of the transformer. The audio waveform diagram is the sound emitted by the transformer during operation, and the infrared video characterizes the temperature performance of each area during the operation of the transformer. That is, the audio waveform diagram and the infrared videos of each area are the responses of the transformer in the operating environment. According to the audio waveform diagram, the areas where the transformer may have faults can be analyzed. Therefore, the search module 202 finds a reference audio waveform diagram from a preset database according to the current data, etc. The reference audio waveform diagram characterizes the sound emitted by the transformer during normal operation when the operating environment of the transformer is relatively similar. The comparison module 203 compares the reference audio waveform diagram with the audio waveform diagram to obtain inconsistent waveform segments in the audio waveform diagram. The stability determination module 204 comprehensively analyzes the inconsistent waveform diagram and the infrared image to determine the operating stability of the transformer. The higher the operating stability, the more stable the overall operation of the transformer, and the smaller the possibility of potential fault locations. On the contrary, the possibility of fault locations is greater. Therefore, finally, the area determination module 205 can analyze the areas where potential faults may occur on the transformer according to the operating stability and the infrared image, and finally achieve more timely discovery of the precise locations where potential faults may occur on the transformer.

[0108] In a possible implementation manner of the embodiment of the present application, when the search module 202 searches for a reference audio waveform diagram from a preset database based on current data, temperature data, and load data, it is specifically used for: Determine the current data, temperature data, and load data at the same time; Calculate the product of the current data and the load data at the same time, and calculate the ratio of the product to the temperature data at the corresponding time; Calculate the first average value of all the ratios, and generate a line chart of the ratios based on all the ratios in the preset historical time period; Calculate the first similarity between the line chart and the line chart in the preset database. The preset database includes multiple audio waveform diagrams, and each audio waveform diagram corresponds to a line chart of ratios; Determine the target line chart with the first similarity reaching the preset similarity threshold from the line charts in the preset database; Calculate the second average value of the ratios in each target line chart, and calculate the first difference between the second average value of each target line chart and the first average value; Determine the matching value of each target line chart based on the first difference and the first similarity, and determine the audio waveform chart corresponding to the target line chart with the highest matching value as the reference audio waveform chart.

[0109] In a possible implementation manner of the embodiment of the present application, when the stability determination module 204 determines the operation stability of the transformer based on the inconsistent waveform segment and the infrared video, it specifically is used for: Determine the first quantity and the total duration of all the inconsistent waveform segments; Determine the target reference waveform segment from the reference audio waveform chart based on the time interval of each inconsistent waveform segment; Determine the distance and the offset angle between the wave peak of each inconsistent waveform segment and the wave peak of the target reference waveform segment; Determine the first outlier of the transformer based on the first quantity, the total duration, the distance, and the deviation angle; Divide the preset historical time period into multiple sub-time periods, and segment the infrared video according to the sub-time periods to obtain sub-video segments corresponding to each sub-time period; Perform gray-scale transformation on each sub-video segment to obtain the corresponding sub-gray-scale video segment, segment the transformer in each sub-gray-scale video segment into multiple regions, and calculate the gray-scale average value and the gray-scale variance of each region in each sub-gray-scale video segment; Determine the second outlier of the transformer based on the gray-scale average value and the gray-scale variance of each region; Determine the operation stability of the transformer based on the first outlier and the second outlier.

[0110] In a possible implementation manner of the embodiment of the present application, when the stability determination module 204 determines the first outlier of the transformer based on the first quantity, the total duration, the distance, and the deviation angle, it specifically is used for: Calculate the deviation score of each inconsistent waveform segment based on the distance and the offset angle; Determine the candidate waveform segments whose deviation scores reach the preset score threshold, and calculate the average value of the deviation scores of all the candidate waveform segments; Calculate the first outlier based on the first quantity, the total duration, the deviation score, and their respective coefficients.

[0111] In a possible implementation manner of the embodiment of the present application, when the area determination module 205 determines the area where the transformer may have potential faults based on the operation stability and the infrared video, it specifically is used for: Determine the maximum gray-scale value of each region in each sub-gray-scale video segment, and calculate the second difference between the maximum gray-scale value and the gray-scale average value of the region; Determine the second quantity of each region where the second difference reaches the preset difference threshold in all the sub-gray-scale video segments; Calculate the proportion of the second quantity to the quantity of all sub-gray video segments; Calculate the third difference between the average gray value of each sub-gray video segment in each region and the preset average value threshold of each region; Calculate the average value of the third differences based on all the third differences of each region; Determine the fault value of each region based on the proportion and the average value of the third differences; Multiply the running stability by the fault value of each region to obtain the probability of a fault occurring in each region; Determine the regions where the probability reaches the preset probability threshold as the regions where the transformer may have potential faults.

[0112] In a possible implementation manner of the embodiment of the present application, when the stability determination module 204 determines the second outlier of the transformer based on the average gray value and the gray variance of each region, it is specifically used for: Determine the target variance of all the average gray values based on the average gray value of all sub-gray video segments of each region; Determine the average value of all the gray variances based on the gray variance of all sub-gray video segments of each region; Calculate the second outlier based on the target variance, the average value of all the gray variances, and their respective corresponding coefficients.

[0113] In a possible implementation manner of the embodiment of the present application, an artificial intelligence-based transformer fault monitoring system 20 further includes: A sending module, configured to send the regions where potential faults may occur to the terminal device of the staff.

[0114] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described artificial intelligence-based transformer fault monitoring system 20 can refer to the corresponding process in the foregoing method embodiment, and will not be elaborated herein.

[0115] In the embodiment of the present application, an electronic device is provided, as Figure 3 shown, Figure 3 The electronic device 30 shown includes: a processor 301 and a memory 303. Among them, the processor 301 and the memory 303 are connected, such as through a bus 302. Optionally, the electronic device 30 may further include a transceiver 304. It should be noted that in actual applications, the transceiver 304 is not limited to one, and the structure of the electronic device 30 does not constitute a limitation to the embodiment of the present application.

[0116] The processor 301 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules, and circuits described in connection with the disclosure of this application. The processor 301 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0117] The bus 302 may include a path for transmitting information between the above components. The bus 302 may be a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus 302 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0118] The memory 303 may be a ROM (Read Only Memory) or other type of static storage device that can store static information and instructions, a RAM (Random Access Memory), or other type of dynamic storage device that can store information and instructions. It may also be an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic storage media, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0119] The memory 303 is used to store the application program code for executing the solution of this application, and is controlled by the processor 301 for execution. The processor 301 is used to execute the application program code stored in the memory 303 to implement the content shown in the foregoing method embodiments.

[0120] Among them, the electronic device includes but is not limited to: mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Tablet Computers), PMPs (Portable Multimedia Players), in-vehicle terminals (such as in-vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. It can also be a server, etc. Figure 3 The illustrated electronic device is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of this application.

[0121] The embodiments of this application provide a computer-readable storage medium, on which a computer program is stored. When it runs on a computer, it enables the computer to execute the corresponding content in the foregoing method embodiments. Compared with the related art, in the embodiments of this application, current data, temperature data, infrared videos, load data, and audio waveform diagrams during operation are obtained, which is convenient for subsequently analyzing the areas where faults may occur on the transformer according to the operation conditions of the transformer during this period. Current data, temperature data, and load data are all key factors characterizing the operating environment of the transformer. The audio waveform diagram is the sound emitted by the transformer during operation, and the infrared video characterizes the temperature performance of each area during the operation of the transformer. That is, the audio waveform diagram and the infrared videos of each area are the responses of the transformer in the operating environment. According to the audio waveform diagram, the areas where the transformer may have faults can be analyzed. Therefore, a reference audio waveform diagram is found from the preset database according to the current data, etc. The reference audio waveform diagram characterizes the sound emitted when the transformer operates normally under a situation similar to the operating environment of the transformer. By comparing the reference audio waveform diagram with the audio waveform diagram, the inconsistent waveform segments in the audio waveform diagram are obtained. By comprehensively analyzing the inconsistent waveform diagram and the infrared image, the operating stability of the transformer can be determined. The higher the operating stability, the more stable the overall operation of the transformer, and the smaller the possibility of potential fault locations. On the contrary, the possibility of fault locations is greater. Therefore, finally, according to the operating stability and the infrared image, the areas where potential faults may occur on the transformer can be analyzed, and ultimately the precise locations where potential faults may occur on the transformer can be discovered more timely.

[0122] It should be understood that although the steps in the flowchart of the accompanying drawings are shown sequentially according to the indication of the arrows, these steps are not necessarily executed sequentially in the order indicated by the arrows. Unless there is a clear indication in this document, there is no strict order restriction for the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or sub-steps or stages of other steps.

[0123] The above are only some embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. An artificial intelligence-based transformer fault monitoring method, characterized in that, Including: Obtaining current data, temperature data, infrared videos, load data, and audio waveform diagrams during operation of the transformer within a preset historical time period; Searching for a reference audio waveform diagram from a preset database based on the current data, temperature data, and load data; Comparing the reference audio waveform diagram with the audio waveform diagram to determine inconsistent waveform segments in the audio waveform diagram; Determining the operation stability of the transformer based on the inconsistent waveform segments and the infrared videos; Determining the area where potential faults may occur in the transformer according to the operation stability and infrared videos.

2. The method for monitoring transformer faults based on artificial intelligence according to claim 1, characterized in that The searching for a reference audio waveform diagram from a preset database based on the current data, temperature data, and load data includes: Determining the current data, temperature data, and load data at the same moment; Calculating the product of the current data and the load data at the same moment, and calculating the ratio of the product to the temperature data at the corresponding moment; Calculating the first average value of all the ratios, and generating a line graph of the ratios based on all the ratios within the preset historical time period; Calculating the first similarity between the line graph and the line graphs in the preset database, where the preset database includes multiple audio waveform diagrams, and each audio waveform diagram corresponds to a line graph of ratios; Determining a target line graph with the first similarity reaching a preset similarity threshold from the line graphs in the preset database; Calculating the second average value of the ratios in each target line graph, and calculating the first difference between the second average value of each target line graph and the first average value; Determining the matching value of each target line graph based on the first difference and the first similarity, and determining the audio waveform diagram corresponding to the target line graph with the highest matching value as the reference audio waveform diagram.

3. The method for monitoring transformer faults based on artificial intelligence according to claim 1, characterized in that, The determining the operation stability of the transformer based on the inconsistent waveform segments and the infrared videos includes: Determining the first quantity and the total duration of all the inconsistent waveform segments; Determining target reference waveform segments from the reference audio waveform diagram based on the time interval of each inconsistent waveform segment; Determining the distance and the offset angle between the peak of each inconsistent waveform segment and the peak of the target reference waveform segment; Determining the first outlier value of the transformer based on the first quantity, the total duration, the distance, and the deviation angle; Dividing the preset historical time period into multiple sub-time periods, and segmenting the infrared videos according to the sub-time periods to obtain sub-video segments corresponding to each sub-time period; Performing grayscale transformation on each sub-video segment to obtain a corresponding sub-grayscale video segment, segmenting the transformer in each sub-grayscale video segment into multiple regions, and calculating the grayscale average value and the grayscale variance of each region in each sub-grayscale video segment; Determining the second outlier value of the transformer based on the grayscale average value and the grayscale variance of each region; Determining the operation stability of the transformer based on the first outlier value and the second outlier value.

4. The method for monitoring transformer faults based on artificial intelligence according to claim 3, characterized in that The determining the first outlier value of the transformer based on the first quantity, the total duration, the distance, and the deviation angle includes: Calculating the deviation score of each inconsistent waveform segment based on the distance and the offset angle; Determine candidate waveform segments whose deviation scores reach a preset score threshold, and calculate the average deviation score of all candidate waveform segments; Calculate the first outlier based on the first quantity, total duration, deviation score, and their respective corresponding coefficients.

5. The method for monitoring transformer faults based on artificial intelligence according to claim 3, characterized in that, Determine the area where the transformer may have potential faults according to the operation stability and infrared video, including: Determine the maximum gray value of each area in each sub-gray video segment, and calculate the second difference between the maximum gray value and the average gray value of the area; Determine the second quantity where the second difference of each area reaches the preset difference threshold in all sub-gray video segments; Calculate the ratio of the second quantity to the number of all sub-gray video segments; Calculate the third difference between the average gray value of each sub-gray video segment of each area and the preset average value threshold of each area; Calculate the average value of the third differences based on all the third differences of each area; Determine the fault value of each area based on the ratio and the average value of the third differences; Multiply the operation stability by the fault value of each area to obtain the probability of a fault occurring in each area; Determine the area where the probability reaches the preset probability threshold as the area where the transformer may have potential faults.

6. The method for monitoring transformer faults based on artificial intelligence according to claim 3, wherein The determining the second outlier of the transformer based on the average gray value and gray variance of each area includes: Determine the target variance of all the average gray values based on the average gray values of all sub-gray video segments of each area; Determine the average value of all the gray variances based on the gray variances of all sub-gray video segments of each area; Calculate the second outlier based on the target variance, the average value of all the gray variances, and their respective corresponding coefficients.

7. A transformer fault monitoring method based on artificial intelligence according to claim 1, characterized in that, The method further includes: Send the area where potential faults may occur to the terminal device of the staff.

8. An artificial intelligence-based transformer fault monitoring system, characterized in that, including: A data acquisition module for acquiring the current data, temperature data, infrared video, load data, and audio waveform diagram during operation of the transformer within a preset historical time period; A search module for searching for a reference audio waveform diagram from a preset database based on the current data, temperature data, and load data; A comparison module for comparing the reference audio waveform diagram with the audio waveform diagram to determine the inconsistent waveform segments in the audio waveform diagram; A stability determination module for determining the operation stability of the transformer based on the inconsistent waveform segments and the infrared video; An area determination module for determining the area where the transformer may have potential faults according to the operation stability and the infrared video.

9. An electronic device, characterized in that, It includes: At least one processor; A memory; At least one application program, where the at least one application program is stored in the memory and is configured to be executed by the at least one processor, and the at least one application program: is used to execute an artificial intelligence-based transformer fault monitoring method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, [[ID=

Images