Extra-high voltage equipment outer surface image feature extraction method, electronic equipment and storage medium
By establishing a parallel time series and feature extraction model in substation equipment, the problem of slow data processing caused by the large number of substation equipment was solved, and fast and accurate thermal status parameter judgment was achieved, thereby improving maintenance efficiency.
Patent Information
- Application Number
- CN202510590629.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-08
- Publication Date
- 2025-09-26
AI Technical Summary
In the existing technology, substation equipment is numerous and complex in structure, which requires operation and maintenance personnel to process a large amount of irrelevant data when analyzing the thermal state parameters of converter transformers, resulting in slow judgment results and affecting maintenance efficiency.
By collecting the operating parameters of substation equipment, a parallel time series is established and input into the feature extraction model of the converter transformer thermal state parameters. Feature correlation calculation and importance test are performed to extract the target operating parameters, remove the interference of low-importance data, and reduce the computational complexity.
The speed of judging the thermal state parameters of the converter transformer is improved, and the maintenance efficiency is improved.
Smart Images
Figure CN120707870A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of converter transformers, and in particular to a method for extracting image features from the outer surface of ultra-high voltage equipment, an electronic device, and a storage medium. Background Art
[0002] At present, to judge the status of the converter transformer, it is necessary to collect the parameters of each device in the substation and obtain the evaluation results by comparing the actual vibration conditions with the predicted vibration conditions under normal operating conditions. Due to the large number of equipment and complex structure in the substation of the hydropower station, its failure mode has the characteristics of complex mechanism, multiple related modules, delayed manifestation, and different situations. The parameters of the equipment are numerous and the operation and maintenance personnel directly analyze all the raw data slowly. In addition, there is a large amount of related data that cannot be directly screened, resulting in slow judgment of the thermal state parameters of the converter transformer, affecting maintenance efficiency. Summary of the Invention
[0003] The main purpose of the present invention is to provide a method for extracting image features from the outer surface of ultra-high voltage equipment, an electronic device and a storage medium, which can remove statistical data interference of low importance and improve maintenance efficiency.
[0004] To achieve the above objectives, the present application provides, in a first aspect, a method for extracting image features from the outer surface of ultra-high voltage equipment, the method comprising:
[0005] Collecting operating parameters of each device in the substation and establishing a parallel time series of the operating parameters of each device in the substation;
[0006] Inputting the operating parameters of each device in the substation into a feature extraction model of thermal state parameters of a converter transformer based on the parallel time series;
[0007] Based on the feature extraction model, feature correlation calculation and importance test of the thermal state parameters of the converter transformer, the target operating parameters of each device in the substation are extracted.
[0008] Optionally, establishing a parallel time series of operating parameters of each device in the substation includes:
[0009] The time of the operating parameters of the converter transformer is divided into unit times, and the time interval of each unit time is the same;
[0010] Sequentially number the unit times according to their time sequence;
[0011] Collecting the operating parameter status of each device in the substation within the unit time;
[0012] Based on the sequence number of the unit time, the operating parameter state quantity of each device is recorded to generate a parallel time series of the operating parameters of each device in the substation.
[0013] Optionally, the feature extraction model of the converter transformer thermal state parameter has a calculation expression as follows:
[0014]
[0015] Among them, S i,j represents the feature extraction model of the thermal state parameters of the converter transformer, i represents the i-th device among the devices in the substation, represents the time series of the jth state quantity corresponding to i, t represents the sampling time corresponding to the state quantity, Indicates the sampling length of the state quantity.
[0016] Optionally, the simplified calculation expression of the feature extraction model of the converter transformer thermal state parameter is:
[0017]
[0018] Wherein, c represents the number of channels for calculation, i.e., for scalar state quantity, c=1, and for vector state quantity, c is equal to the data dimension.
[0019] Optionally, the extracting target operating parameters of each device in the substation based on the feature extraction model, feature correlation calculation, and importance test of the thermal state parameters of the converter transformer includes:
[0020] Based on the feature extraction model of the thermal state parameters of the converter transformer, the statistics of the operating parameters of each device in the substation are extracted. The statistics of the operating parameters of each device include the skewness of the sequence, the length of the sequence, the kurtosis of the sequence, the quantile of the empirical distribution function of the sequence, and the bucket entropy of the sequence. The calculation expression is:
[0021]
[0022] Among them, S represents the state vector under a single channel, s v represents the vth component in the vector, ɑ represents the sequence number, n t represents the length of the sequence, skewness(S) represents the skewness of the sequence S, kurtosis(S) represents the kurtosis of the sequence, std(S) represents the standard deviation of the sequence, q represents the empirical distribution function of the sequence, Q q (S) represents the quantile of the empirical distribution function q of the sequence, E m (S) represents the bucket entropy of the sequence, m represents the number of intervals, p represents the importance of the statistical feature correlation of the operating parameters of each device in the substation in the sequence interval, and pk represents the probability corresponding to the kth interval.
[0023] Optionally, the extracting target operating parameters of each device in the substation based on the feature extraction model, feature correlation calculation, and importance test of the thermal state parameters of the converter transformer further includes:
[0024] Calculate the statistical feature correlation between the converter transformer state and the operating parameters of each substation device, and test the importance of the statistical feature correlation; wherein a low p-value indicates a low statistical feature correlation between the operating parameters of each substation device, and a high p-value indicates a high statistical feature correlation between the operating parameters of each substation device;
[0025] The statistical quantities of the operating parameters of each device in the substation are selected based on the size of the p value.
[0026] Optionally, the calculating of the statistical characteristic correlation between the converter transformer state and the operating parameters of each substation device, and testing the importance thereof, includes:
[0027] According to the different feature types and sample categories, Fisher's exact test, KS test, and Kendall's rank test are used to obtain the p value to determine whether the hypothesis is true. The statistics of the operating parameters of each device can be expressed as n, and the corresponding hypothesis test results can be expressed as The calculation expression of the Fisher's exact test is:
[0028]
[0029] The calculation expression of the KS test is:
[0030]
[0031] The calculation expression of the Kendall rank test is:
[0032]
[0033] Among them, H0 assumes that the feature is irrelevant to the sample category prediction, H1 assumes that the feature is relevant to the sample category prediction, the state numbers are m and n, their variables are X and Y, i and j represent the state numbers, N represents the total number of samples, R i represents the number of samples of X=i, C j represents the number of samples of Y=j, a ij represents the number of samples with X=i, Y=j, F represents the cumulative distribution function, and Y i represents the variable numbered i in the ordered variable sequence, N represents the total number of samples, P1 and P2 represent the ordered independent variable sequences corresponding to different dependent variable categories, and d ΔIndicates distance calculation, and N indicates the total number of samples.
[0034] Optionally, the extracting target operating parameters of each device in the substation based on the feature extraction model, feature correlation calculation, and importance test of the thermal state parameters of the converter transformer further includes:
[0035] The p-value sequence is sorted, a threshold is calculated according to a set value of the false discovery rate, statistics with p-values lower than the threshold are retained, and the target operating parameters of each device in the substation are extracted.
[0036] A second aspect of the present application provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the first aspect and any possible implementation thereof.
[0037] A third aspect of the present application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the method described in the first aspect.
[0038] The present application provides a method, electronic device, and storage medium for extracting features from the external surface image of ultra-high voltage equipment. The method collects operating parameters of various devices in a substation and establishes a parallel time series of the operating parameters of the various devices in the substation; inputs the operating parameters of the various devices in the substation into a feature extraction model for thermal state parameters of a converter transformer based on the parallel time series; extracts target operating parameters of the various devices in the substation based on the feature extraction model, feature correlation calculation, and importance test of the thermal state parameters of the converter transformer; and removes interference from low-importance data through multi-dimensional statistical analysis and feature importance test, reduces the complexity of the calculation and analysis process, improves the calculation efficiency of model data, and accelerates the judgment speed of the thermal state parameters of the converter transformer, thereby improving maintenance efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0040] in:
[0041] Figure 1 A schematic flow chart of a method for extracting image features from the outer surface of ultra-high voltage equipment provided in an embodiment of the present application;
[0042] Figure 2 A schematic diagram of a parallel time series flow chart for extracting image features from the outer surface of ultra-high voltage substation equipment provided in an embodiment of the present application;
[0043] Figure 3 A schematic diagram of a parallel time series feature extraction process of a method for extracting image features from the outer surface of ultra-high voltage substation equipment provided in an embodiment of the present application;
[0044] Figure 4 A schematic diagram showing the distribution of four evaluation levels for different equipment types in an embodiment of a method for extracting image features from the outer surface of ultra-high voltage substation equipment provided in an embodiment of the present application;
[0045] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0046] In order to enable those skilled in the art to better understand the present invention, 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 those skilled in the art without creative work are within the scope of protection of this application.
[0047] The terms "first," "second," and the like in the specification and claims of this application and the accompanying drawings are used to distinguish between different objects, not to describe a particular order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or apparatus.
[0048] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0049] The embodiments of the present application are described below in conjunction with the drawings in the embodiments of the present application.
[0050] See also Figure 1, is a flow chart of a method for extracting image features from the outer surface of ultra-high voltage equipment provided in an embodiment of the present application, such as Figure 1 As shown, the method includes:
[0051] 101. Collect operating parameters of each device in the substation and establish a parallel time series of the operating parameters of each device in the substation.
[0052] The execution subject of the method in the embodiment of the present application can be a device for extracting image features from the outer surface of ultra-high voltage equipment. In practical applications, it can specifically be an electronic device.
[0053] Specifically in this embodiment, the equipment of the substation can be divided into an electromagnetic unit, a mechanical unit, a control unit, and an integrated unit, wherein the electromagnetic unit is the electromagnetic quantity corresponding to the primary equipment, including the partial discharge of the generator and transformer, the air gap of the generator set, the magnetic field strength, and the core grounding current; the mechanical unit is the state quantity corresponding to the mechanical part of the turbine and the generator, including the vibration of the frame and support cover, the swing of the upper guide and water guide, the vibration of the stator core, the water pressure pulsation of the turbine and the degree of cavitation, and the water pressure conditions in different parts; the control unit is the monitoring quantity corresponding to the control system, including the head, unit speed, guide vane opening, terminal voltage, stator current, excitation current, active power, reactive power, system-related control parameters and given, and switch quantity information; the integrated unit corresponds to the unit-related temperature quantities such as the upper guide bearing temperature and stator temperature, as well as information such as the oil level and pressure of auxiliary equipment.
[0054] Currently, the parameters of various equipment in the substation are collected through sensors, and then the sampled signals are collected and sent to the converter transformer to monitor the equipment in the substation. For high-sampling-rate signals such as partial discharge, the storage space required for all the original data of the equipment's operating parameters is too large. Therefore, the converter transformer system generally only retains data records that exceed the limit value. That is, when a certain state quantity exceeds the limit at a certain moment, the system will record a fixed length of data before and after that moment; in addition, in order to obtain the normal state quantity distribution under different working conditions, some normal data can be randomly selected for recording.
[0055] Figure 2 A schematic diagram of establishing a parallel time series flow in a method for extracting image features from the outer surface of ultra-high voltage substation equipment provided in one embodiment of the present application.
[0056] In an optional embodiment, the step of establishing a parallel time series of operating parameters of each device in the substation includes:
[0057] 021. Divide the time of the operating parameters of the converter transformer into unit times, and the time interval of each unit time is the same;
[0058] 022. Sequentially number the above-mentioned unit times according to their chronological order;
[0059] 023. Collect the operating parameter status of each device in the above-mentioned substation within the above-mentioned unit time;
[0060] 024. Based on the above-mentioned sequence number of the unit time, the operating parameter state quantity of each device is recorded to generate a parallel time series of the operating parameters of each device in the above-mentioned substation.
[0061] Specifically, the time of the operating parameters of the converter transformer is divided into unit times, and the time interval of each unit time is the same; since the sequence scales of different state quantities, that is, the sampling frequencies, vary greatly, and there are correlations between different devices and different state quantities, in order to reduce the error of collected data, the time interval of each unit time is made the same.
[0062] The unit times are serially numbered according to the order of time; this is to facilitate the organization and calculation of the unit times.
[0063] The operating parameter status quantities of each device in the substation are collected within a unit time; the operating parameter status quantities of each device in the substation need to be collected at the same time within a unit time.
[0064] Based on the time unit sequence number, the operating parameter status of each device is recorded to generate a parallel time series of the operating parameters of each substation device. The time units corresponding to the sequence numbers and the operating parameter status of the corresponding devices are sorted to form a data set of the operating parameter status of each substation device for a certain time unit, that is, a parallel time series of the operating parameters of each substation device.
[0065] 102. Input the operating parameters of each device in the above-mentioned substation into the feature extraction model of the thermal state parameters of the converter transformer based on the above-mentioned parallel time series.
[0066] Specifically, in this implementation, by generating a parallel time series of the operating parameters of each device in the substation, it is convenient for operation and maintenance personnel to conduct statistics on the operating parameters of each device in the substation. At present, operation and maintenance personnel generally use the operating parameters of each device. Due to the large number of operating parameters of the equipment, the operation and maintenance personnel directly analyze all the original data slowly and cannot directly perform screening, resulting in slow judgment results on the thermal state parameters of the converter transformer, affecting maintenance efficiency. By extracting interrelated data from the features of the thermal state parameters of the converter transformer, it is beneficial to reduce the workload of analysis and statistics, quickly obtain equipment operating condition information, and judge whether the equipment needs to be replaced or repaired.
[0067] In an optional embodiment, the calculation expression of the feature extraction model of the thermal state parameter of the converter transformer is:
[0068]
[0069] Among them, S i,j represents the feature extraction model of the thermal state parameters of the converter transformer, i represents the i-th device among the devices in the substation, represents the time series of the jth state quantity corresponding to i, t represents the sampling time corresponding to the state quantity, Indicates the sampling length of the state quantity.
[0070] Further optionally, the simplified calculation expression of the feature extraction model of the converter transformer thermal state parameter is:
[0071]
[0072] Wherein, c represents the number of channels for calculation, i.e., for scalar state quantity, c=1, and for vector state quantity, c is equal to the data dimension.
[0073] 103. Based on the feature extraction model, feature correlation calculation and importance test of the above-mentioned converter transformer thermal state parameters, extract the target operating parameters of each equipment in the above-mentioned substation.
[0074] Specifically in this implementation, the statistics of the operating parameters of each device may also include the maximum value, minimum value, mean, variance and standard deviation. The maximum value, minimum value, mean, variance and standard deviation can be calculated by existing mathematical formulas to obtain their corresponding statistics; in the embodiment of the present application, the operating parameters of each device can be mainly extracted by performing calculation and analysis on the statistics of the operating parameters of each device in multiple dimensions, through feature correlation and testing the importance of the feature correlation; among them, the feature correlation of the kurtosis of the sequence, the standard deviation of the sequence, the quantile of the empirical distribution function q of the sequence and the bucket entropy of the sequence can be involved, and the importance of the feature correlation can be tested.
[0075] In an optional implementation, the above step 103 includes:
[0076] Based on the feature extraction model of the thermal state parameters of the converter transformer, the statistics of the operating parameters of the above-mentioned substation equipment are extracted. The statistics of the operating parameters of the above-mentioned equipment include the skewness of the sequence, the length of the sequence, the kurtosis of the sequence, the quantile of the empirical distribution function of the sequence, and the bucket entropy of the sequence. The calculation expression is:
[0077]
[0078] Among them, S represents the state vector under a single channel, s v represents the vth component in the vector, ɑ represents the sequence number, n trepresents the length of the sequence, skewness(S) represents the skewness of the sequence S, kurtosis(S) represents the kurtosis of the sequence, std(S) represents the standard deviation of the sequence, q represents the empirical distribution function of the sequence, Q q (S) represents the quantile of the empirical distribution function q of the sequence, E m (S) represents the bucket entropy of the sequence, m represents the number of intervals, p represents the importance of the statistical feature correlation of the operating parameters of each device in the substation in the sequence interval, and p k represents the probability corresponding to the kth interval.
[0079] Furthermore, the above step 103 further includes:
[0080] Calculate the statistical feature correlation between the converter transformer status and the operating parameters of each substation device, and test its importance; a low p-value indicates a low statistical feature correlation between the operating parameters of each substation device, and a high p-value indicates a high statistical feature correlation between the operating parameters of each substation device;
[0081] The statistical quantities of the operating parameters of each device in the substation are selected based on the size of the above p-value.
[0082] Specifically, in this implementation, through multi-dimensional statistics of the operating parameters of each device, the statistical characteristic correlation between the status of the converter transformer and the operating parameters of each device in the substation is calculated, and the importance of each is tested. This is conducive to removing the interference of statistical data with low importance, reducing the complexity of the calculation and analysis process, improving the calculation efficiency of model data, accelerating the speed of judging the results of the thermal state parameters of the converter transformer, and improving maintenance efficiency.
[0083] In an optional embodiment, the above-mentioned calculation of the statistical characteristic correlation between the converter transformer state and the operating parameters of each substation device and the testing of their importance includes:
[0084] According to the different feature types and sample categories, Fisher's exact test, KS test, and Kendall's rank test are used to obtain the p value to determine whether the hypothesis is true. The statistics of the operating parameters of the above-mentioned equipment can be expressed as n, and the corresponding hypothesis test results can be expressed as The calculation expression of the above Fisher's exact test is:
[0085]
[0086] The calculation expression of the above KS test is:
[0087]
[0088] The calculation expression of the above Kendall rank test is:
[0089]
[0090] Among them, H0 hypothesis states that this feature has nothing to do with the sample category prediction. H 1 Assume that the feature is related to the sample category prediction, the state numbers are m and n, their variables are X and Y, i and j represent the state numbers, N represents the total number of samples, R i represents the number of samples of X=i, C j represents the number of samples of Y=j, a ij represents the number of samples with X=i, Y=j, F represents the cumulative distribution function, and Y i represents the variable numbered i in the ordered variable sequence, N represents the total number of samples, P1 and P2 represent the ordered independent variable sequences corresponding to different dependent variable categories, and d Δ Indicates distance calculation, and N indicates the total number of samples.
[0091] Optionally, the above step 103 further includes:
[0092] The p-value sequence is sorted, a threshold is calculated according to a set value of the false discovery rate, and statistics with p-values lower than the threshold are retained to complete the extraction of target operating parameters of each device in the substation.
[0093] Specifically, in order to control the false discovery rate and determine the statistical significance threshold in the embodiment of the present application, the following steps are taken:
[0094] Sort the P-value sequence: First, sort the P-values obtained from all hypothesis tests from smallest to largest. This step is to prepare for the subsequent calculation of the threshold;
[0095] Calculate the threshold: Calculate the threshold r according to the set value of the false discovery rate φ , and its calculation formula is:
[0096]
[0097] Among them, r φ represents the threshold, φ represents the hypothesis number after sorting by p value, q represents the false discovery rate setting value, n φ Represents the total number of hypotheses. This threshold is used to determine which hypothesis test results are statistically significant in multiple comparisons.
[0098] Determining the False Discovery Rate (FDR): The FDR is determined by comparing and analyzing the same type of monitoring data from the same device in both non-interference and interference-exposed conditions. Specifically, data exceeding the data error threshold is considered erroneous data, and the error rate of the statistic for this erroneous data is calculated to determine the FDR.
[0099] Retain statistics with P values below the threshold: Finally, retain those statistics with P values below the calculated threshold. These statistics are considered significant under the premise of controlling the false discovery rate and can be used to complete the extraction of operating parameters of various substation equipment.
[0100] The above steps can remove the interference of statistical data with low importance, reduce the complexity of the calculation and analysis process, improve the efficiency of model data calculation, speed up the judgment of the thermal state parameters of the converter transformer, and improve the maintenance efficiency.
[0101] Figure 3 A schematic diagram of a parallel time series feature extraction process for a method for extracting image features from the outer surface of ultra-high voltage substation equipment provided in one embodiment of the present application.
[0102] Figure 4 A schematic diagram of the distribution of four evaluation levels for different equipment types in an embodiment of a method for extracting image features from the outer surface of ultra-high voltage substation equipment provided in one embodiment of the present application.
[0103] Reference Figure 3 and 4 This is another embodiment of the present invention. This embodiment differs from the first embodiment in that it provides an experimental verification of a method for extracting image features from the exterior surface of ultra-high voltage substation equipment, demonstrating the technical effectiveness of this method. This embodiment compares conventional technical solutions with the method of this application, using scientifically proven methods to compare the test results and verify the effectiveness of this method.
[0104] The operating data of a hydropower station in a certain place recorded by the computer monitoring system of a hydropower enterprise were selected for analysis and confirmation.
[0105] Data collected from different devices and sensors requires a unified monitoring and diagnostic time unit; this method uses days as the unit. Daily data from each device and monitoring unit is collected, including relevant parameters such as active power, reactive power, water supply pressure and direction, component swing, stator temperature, cold air temperature, hot air temperature, upper guide bearing pad temperature and oil sump level, thrust bearing pad temperature and oil sump level. Based on relevant thresholds and trend analysis, the system initially determines the status of each device, which is then further confirmed and verified by operations and maintenance personnel.
[0106] For the equipment-level status evaluation problem, a single case is a different monitoring data record corresponding to a certain device on a single day. According to statistics, the above dataset contains a total of 2816 cases, corresponding to a duration of 176 days. For each case, it is divided into four levels: normal, caution, abnormal, and severe according to the type of equipment and operating conditions. The number of each category is 1968, 533, 533, and 75 respectively. The distribution of the four evaluation levels for different equipment types is shown in the figure below. Figure 4 shown.
[0107] Table 1 is a feature extraction verification record table of four levels in different device types provided in this application.
[0108] Statistical feature count Accurate number of feature extraction Verification test accuracy Normal state 1968 1869 94.9％ Attention Status 533 489 91.7％ Abnormal status 533 516 96.8％ Severe condition 75 74 98.6％
[0109] Table 1
[0110] It can be seen that the feature extraction models of normal state and attention state extract the statistical feature correlation of the operating parameters of each device in the substation, but the accuracy is not obvious, because the parameter variation range of these two states is small, and the parameter correlation interference between each device is relatively strong; and as the state level increases, the feature extraction model extracts the statistical feature correlation of the operating parameters of each device in the substation, and after the importance test, it can be seen that the accuracy of the feature extraction model gradually increases with the abnormality of the equipment parameters, which is conducive to removing the interference of statistical data with low importance, reducing the complexity of the calculation and analysis process, improving the calculation efficiency of model data, accelerating the judgment speed of the thermal state parameters of the converter transformer, and improving the maintenance efficiency.
[0111] Based on the description of the aforementioned method embodiment, an embodiment of the present application further provides an electronic device.
[0112] See Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 5 As shown, the electronic device 500 includes a processor 501 and a memory 502. The memory 502 stores a computer program. When the computer program is executed by the processor 501, the following operations are performed: Figure 1 The electronic device 500 may further include an input / output device, etc. In a specific embodiment, the electronic device may be a terminal device, etc.
[0113] In one embodiment, a computer-readable storage medium is further provided. The computer-readable storage medium stores a computer program. When the computer program is executed by the processor 501, the processor 501 executes any step in the above method embodiment.
[0114] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
[0115] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0116] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.
Claims
1. A method for extracting image features from the outer surface of ultra-high voltage equipment, characterized in that: include: Collecting operating parameters of each device in the substation and establishing a parallel time series of the operating parameters of each device in the substation; Inputting the operating parameters of each device in the substation into a feature extraction model of thermal state parameters of a converter transformer based on the parallel time series; Based on the feature extraction model, feature correlation calculation and importance test of the thermal state parameters of the converter transformer, the target operating parameters of each device in the substation are extracted.
2. The method for extracting image features from the outer surface of ultra-high voltage equipment according to claim 1, characterized in that: The establishing of a parallel time series of operating parameters of each device in the substation includes: The time of the operating parameters of the converter transformer is divided into unit times, and the time interval of each unit time is the same; Sequentially number the unit times according to their time sequence; Collecting the operating parameter status of each device in the substation within the unit time; Based on the sequence number of the unit time, the operating parameter state quantity of each device is recorded to generate a parallel time series of the operating parameters of each device in the substation.
3. The method for extracting image features from the outer surface of ultra-high voltage equipment according to claim 1, characterized in that: The calculation expression of the feature extraction model of the converter transformer thermal state parameter is: Among them, S i,j represents the feature extraction model of the thermal state parameters of the converter transformer, i represents the i-th device among the devices in the substation, represents the time series of the jth state quantity corresponding to i, t represents the sampling time corresponding to the state quantity, Indicates the sampling length of the state quantity.
4. The method for extracting image features from the outer surface of ultra-high voltage equipment according to claim 3, characterized in that: The simplified calculation expression of the feature extraction model of the converter transformer thermal state parameter is: Wherein, c represents the number of channels for calculation, i.e., for scalar state quantity, c=1, and for vector state quantity, c is equal to the data dimension.
5. The method for extracting image features from the outer surface of ultra-high voltage equipment according to claim 1, characterized in that: The target operating parameters of each device in the substation are extracted based on the feature extraction model, feature correlation calculation and importance test of the thermal state parameters of the converter transformer, including: Based on the feature extraction model of the thermal state parameters of the converter transformer, the statistics of the operating parameters of each device in the substation are extracted. The statistics of the operating parameters of each device include the skewness of the sequence, the length of the sequence, the kurtosis of the sequence, the quantile of the empirical distribution function of the sequence, and the bucket entropy of the sequence. The calculation expression is: Among them, S represents the state vector under a single channel, s v represents the vth component in the vector, ɑ represents the sequence number, n t represents the length of the sequence, skewness(S) represents the skewness of the sequence S, kurtosis(S) represents the kurtosis of the sequence, std(S) represents the standard deviation of the sequence, q represents the empirical distribution function of the sequence, Q q (S) represents the quantile of the empirical distribution function q of the sequence, E m (S) represents the bucket entropy of the sequence, m represents the number of intervals, p represents the importance of the statistical feature correlation of the operating parameters of each device in the substation in the sequence interval, and p k represents the probability corresponding to the kth interval.
6. The method for extracting image features from the outer surface of UHV equipment according to claim 5, characterized in that: The feature extraction model, feature correlation calculation and importance test based on the thermal state parameters of the converter transformer are used to extract the target operating parameters of each device in the substation, and further include: Calculate the statistical feature correlation between the converter transformer state and the operating parameters of each substation device, and test the importance of the statistical feature correlation; wherein a low p-value indicates a low statistical feature correlation between the operating parameters of each substation device, and a high p-value indicates a high statistical feature correlation between the operating parameters of each substation device; The statistical quantities of the operating parameters of each device in the substation are selected based on the size of the p value.
7. The method for extracting image features from the outer surface of ultra-high voltage equipment according to claim 6, characterized in that: The calculation of the statistical characteristic correlation between the converter transformer state and the operating parameters of each substation device and the testing of their importance includes: According to the different feature types and sample categories, Fisher's exact test, KS test, and Kendall's rank test are used to obtain the p value to determine whether the hypothesis is true. The statistics of the operating parameters of each device can be expressed as n, and the corresponding hypothesis test results can be expressed as The calculation expression of the Fisher's exact test is: The calculation expression of the KS test is: The calculation expression of the Kendall rank test is: Among them, H0 assumes that the feature is irrelevant to the sample category prediction, H1 assumes that the feature is relevant to the sample category prediction, the state numbers are m and n, their variables are X and Y, i and j represent the state numbers, N represents the total number of samples, R i represents the number of samples of X=i, C j represents the number of samples of Y=j, a ij represents the number of samples with X=i, Y=j, F represents the cumulative distribution function, and Y i represents the variable numbered i in the ordered variable sequence, N represents the total number of samples, P1 and P2 represent the ordered independent variable sequences corresponding to different dependent variable categories, and d Δ Indicates distance calculation, and N indicates the total number of samples.
8. The method for extracting image features from the outer surface of ultra-high voltage equipment according to claim 7, characterized in that: The feature extraction model, feature correlation calculation and importance test based on the thermal state parameters of the converter transformer are used to extract the target operating parameters of each device in the substation, and further include: The p-value sequence is sorted, a threshold is calculated according to a set value of the false discovery rate, statistics with p-values lower than the threshold are retained, and the target operating parameters of each device in the substation are extracted.
9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method according to any one of claims 1 to 8.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 8.