A method for selecting sensitive areas of winding vibration on the surface of power transformer oil tank

Abstract

The present invention belongs to the technical field of power transformer oil tank management, and specifically is a method for selecting a sensitive area of ​​winding vibration on the surface of a power transformer oil tank, comprising measurement point layout rationality analysis, vibration signal multi-point acquisition, vibration signal preprocessing, feature extraction analysis and sensitive area selection output; the present invention performs layout rationality analysis through a measurement point layout scanning module to determine whether to generate a layout warning signal, and adjusts the layout of vibration measurement points when generating a layout warning signal, which is beneficial to ensuring the accuracy of the finally selected sensitive area; when no layout warning signal is generated, the sensitive area of ​​winding vibration on the surface of the power transformer oil tank is determined through vibration signal acquisition, preprocessing, feature extraction analysis and comprehensive sensitivity evaluation, ensuring that the selected area can accurately reflect the vibration characteristics and state of the transformer winding, and significantly improving the monitoring accuracy.

Description

Technical Field

[0001] The invention relates to the technical field of power transformer oil tank management, in particular to a method for selecting a sensitive area of ​​winding vibration on the surface of a power transformer oil tank. Background Art

[0002] Power transformers are key equipment for power transmission and distribution. Monitoring their operating status is crucial to ensuring the stability and safety of the power system. Winding vibration is an important indicator of changes in the internal state of the transformer. By monitoring the vibration signal on the surface of the tank, the health of the winding can be indirectly evaluated.

[0003] However, the vibration signal characteristics at different locations on the surface of the fuel tank are different, making it difficult to effectively and accurately select sensitive areas, which is not conducive to ensuring monitoring accuracy. In addition, before selecting sensitive areas, it is not possible to reasonably judge the rationality of the layout of vibration measurement points and issue timely warnings, which is not conducive to managers adjusting the measurement point layout, and the degree of intelligence is low.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a method for selecting sensitive areas of winding vibration on the surface of an oil tank of a power transformer, which solves the problems that the prior art is difficult to effectively and accurately select sensitive areas, which is not conducive to ensuring monitoring accuracy, and before selecting sensitive areas, it is impossible to reasonably judge the rationality of the layout of vibration measuring points and provide timely warnings, and has a low degree of intelligence.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for selecting a sensitive area of ​​winding vibration on the surface of a power transformer oil tank comprises the following steps:

[0008] Step 1: The measuring point layout scanning module collects the surface image of the power transformer oil tank, obtains the positions of all vibration measuring points based on the surface image, determines whether to generate a layout warning signal through layout rationality analysis, and sends the layout warning signal to the supervision terminal through the processor;

[0009] Step 2: If no layout warning signal is generated, a vibration signal is synchronously collected at each vibration measuring point using a high-precision vibration sensor through a vibration signal collection module, and the collected vibration signal is sent to a signal preprocessing module through a processor;

[0010] Step 3: The signal preprocessing module preprocesses the collected vibration signal, wherein the preprocessing operation includes filtering, denoising and normalization steps, and sends the preprocessed vibration signal to the feature extraction and analysis module through the processor;

[0011] Step 4: The feature extraction and analysis module analyzes the preprocessed vibration signal and sends the analysis result to the sensitive area evaluation and selection module through the processor;

[0012] Step 5. The sensitive area assessment selection module constructs a sensitive area assessment model, calculates a comprehensive sensitivity score for each vibration measurement point through the sensitive area assessment model and based on the analysis results of the feature extraction and analysis module, selects the location of the vibration measurement point with the highest score as the sensitive area according to the comprehensive sensitivity score result, and sends the selected sensitive area to the supervision terminal through the processor.

[0013] Furthermore, in step 1, the specific analysis method of layout rationality analysis is as follows:

[0014] Based on the surface image of the power transformer oil tank, the high-impact structure and high-impact accessories on the oil tank are identified; the high-impact structure includes reinforcing ribs and support frames, and the high-impact accessories include oil inlet and outlet ports, cooler interfaces and valves; the high-impact structure and high-impact accessories are marked as reference objects, and the vibration measurement points distributed on the corresponding reference objects are marked as unreasonable layout measurement points;

[0015] If there is no vibration measuring point on the corresponding reference object, the shortest distance between the corresponding reference object and the corresponding vibration measuring point is collected and marked as a reference distance detection value, and the reference distance detection value is numerically compared with the corresponding preset reference distance detection threshold. If the reference distance detection value does not exceed the corresponding preset reference distance detection threshold, the corresponding vibration measuring point is marked as an unreasonable layout measuring point;

[0016] If there are unreasonable layout measurement points on the surface of the power transformer oil tank, a layout warning signal will be generated.

[0017] If there are no unreasonable layout measurement points on the surface of the power transformer oil tank, several detection areas are set on the surface of the power transformer oil tank. Through analysis, it is determined whether the corresponding detection area is a non-optimal layout area. If there is a non-optimal layout area on the surface of the power transformer oil tank, a layout warning signal is generated.

[0018] Furthermore, the analysis and judgment method of the layout of non-optimal areas is as follows:

[0019] The number of vibration measurement points in the corresponding detection area is collected and marked as the measurement point number inspection value, and the measurement point number inspection value is compared with the corresponding preset measurement point number inspection value range. If the measurement point number inspection value is not within the corresponding preset measurement point number inspection value range, the corresponding detection area is marked as a non-optimal layout area.

[0020] Furthermore, if the number of measurement points is within the corresponding preset number of measurement points, the distance between any two groups of vibration measurement points in the corresponding detection area is marked as a point distance detection value, and the point distance detection value is numerically compared with the preset point distance detection threshold. If the point distance detection value does not exceed the preset point distance detection threshold, the corresponding point distance detection value is marked as a point distance abnormal value.

[0021] The number of point distance anomaly values ​​in the corresponding detection area is obtained and the ratio thereof is calculated with the number of point distance detection values ​​to obtain the point distance anomaly value, and the difference between the point distance anomaly value and the preset point distance detection threshold is marked as the point distance amplitude detection value, the mean of all point distance anomaly values ​​is calculated to obtain the point distance amplitude value, and the point distance amplitude detection value with the largest value is marked as the point distance amplitude analysis value;

[0022] The regional distribution evaluation value is obtained by numerically calculating the point distance anomaly value, the point distance amplitude value and the point distance amplitude analysis value, and the regional distribution evaluation value is numerically compared with the corresponding preset regional distribution evaluation threshold. If the regional distribution evaluation value exceeds the preset regional distribution evaluation threshold, the corresponding detection area will be marked as a non-optimal layout area.

[0023] Furthermore, the processor is communicatively connected to a sensitive area reselection warning module, which is used to re-analyze the selection urgency of the sensitive area of ​​the winding vibration on the surface of the power transformer oil tank, determine whether to generate a reselection warning signal through analysis, and send the reselection warning signal to the supervision terminal via the processor. The supervision terminal issues a corresponding warning when receiving the reselection warning signal.

[0024] Furthermore, the specific analysis process of the sensitive area reselection warning module includes:

[0025] The selection moment of the last sensitive area where the winding on the surface of the power transformer oil tank vibrates is obtained and marked as the neighboring selection moment, and the interval between the neighboring selection moment and the current moment is marked as the neighboring selection interval value. The neighboring selection interval value is numerically compared with the preset neighboring selection interval threshold. If the neighboring selection interval value exceeds the preset neighboring selection interval threshold, a reselection warning signal is generated.

[0026] Furthermore, if the neighbor selection interval value does not exceed the preset neighbor selection interval threshold, the neighbor selection shape measurement value and the neighbor selection impact value are obtained through analysis, and the ratio of the neighbor selection interval value to the preset neighbor selection interval threshold is marked as the neighbor selection interval occupancy value;

[0027] The re-election warning coefficient is obtained by numerically calculating the neighboring separation occupancy value, the neighboring separation shape measurement value and the neighboring separation impact value, and the re-election warning coefficient is numerically compared with the preset re-election warning coefficient threshold. If the re-election warning coefficient exceeds the preset re-election warning coefficient threshold, a re-election warning signal is generated.

[0028] Further, the analysis and acquisition method of the adjacent shape measurement value is as follows:

[0029] The deformation value of several detection points on the surface of the oil tank of the power transformer within the interval between the adjacent selected time and the current time is collected by the deformation sensor, and the deformation value of the corresponding detection point is numerically compared with the preset deformation threshold value. If the deformation value exceeds the preset deformation threshold value, the corresponding detection point is marked as a variation point;

[0030] The number of variation points is obtained and the ratio is calculated with the total number of detection points to obtain the variation detection value, and the deformation values ​​of all detection points are averaged to obtain the deformation analysis value, and the deformation value with the largest value is marked as the deformation amplitude value. The adjacent shape measurement value is obtained by numerically calculating the variation detection value, deformation analysis value and deformation amplitude value.

[0031] Further, the analysis and acquisition method of the adjacent impact value is as follows:

[0032] When the surface of the oil tank of the power transformer is impacted by an object, the impact force of the collision impact process is collected through a pressure sensor, the maximum impact force, the average impact force, and the duration of the collision impact in the collision impact process are obtained and marked as the collision condition value, the maximum impact force, the average impact force and the collision condition value are numerically calculated to obtain the collision damage detection value, the collision damage detection value is numerically compared with the preset collision damage detection threshold, if the collision damage detection value exceeds the preset collision damage detection threshold, the corresponding collision impact process is marked as an abnormal collision process;

[0033] The total number of times the surface of the power transformer oil tank is impacted by objects during the interval between the adjacent selected moment and the current moment is obtained and marked as the total frequency value of damage, and the number of abnormal collision processes on the surface of the power transformer oil tank during the interval between the adjacent selected moment and the current moment is obtained and marked as the total frequency value of abnormal collisions. The total frequency value of damage and the total frequency value of abnormal collisions are weighted and summed to obtain the adjacent impact value.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1. In the present invention, a layout rationality analysis is performed through a measuring point layout scanning module to determine whether to generate a layout warning signal. When a layout warning signal is generated, the layout of the vibration measuring points is adjusted to ensure the accuracy of the finally selected sensitive area. When a layout warning signal is not generated, the sensitive area of ​​the winding vibration on the surface of the power transformer oil tank is determined through vibration signal collection, preprocessing, feature extraction analysis and comprehensive sensitivity evaluation to ensure that the selected area can accurately reflect the vibration characteristics and state of the transformer winding, thereby significantly improving the accuracy of monitoring;

[0036] 2. In the present invention, the selection urgency of the sensitive area of ​​the winding vibration on the surface of the power transformer oil tank is re-analyzed through the sensitive area reselection warning module, and the analysis is used to determine whether to generate a reselection warning signal. When the reselection warning signal is generated, the management personnel are reminded to reselect the sensitive area of ​​the winding vibration on the surface of the power transformer oil tank in time, so as to ensure the subsequent monitoring effect, reduce the management difficulty, and have a high degree of intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0038] Figure 1 This is a flow chart of the method of Embodiment 1 of the present invention;

[0039] Figure 2 This is a system block diagram of Embodiment 1 of the present invention;

[0040] Figure 3 This is a system block diagram of Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] Embodiment 1: Figure 1-2 As shown, the present invention proposes a method for selecting a sensitive area of ​​winding vibration on the surface of a power transformer oil tank, comprising the following steps:

[0043] Step 1: The scanning module collects the surface image of the oil tank of the power transformer, obtains the positions of all vibration measuring points based on the surface image, and determines whether to generate a layout warning signal through layout rationality analysis. The layout warning signal is sent to the supervision terminal through the processor. When the supervision terminal receives the layout warning signal, it issues a corresponding warning to remind the management personnel to adjust the layout of the vibration measuring points, which is conducive to ensuring the accuracy of the sensitive area finally selected; the specific analysis method of the layout rationality analysis is as follows:

[0044] Based on the surface image of the power transformer tank, high-impact structures and high-impact accessories on the tank are identified; among them, high-impact structures include structures such as reinforcing ribs and support frames (these areas are prone to affect the transmission of vibration signals and should be avoided as vibration measurement points as much as possible), and high-impact accessories include accessories such as oil inlets and outlets, cooler interfaces and valves (these areas are not ideal measurement point locations and are prone to introduce additional vibration noise); high-impact structures and high-impact accessories are marked as reference objects, and vibration measurement points distributed on the corresponding reference objects are marked as measurement points with unreasonable layout;

[0045] If there is no vibration measuring point on the corresponding reference object, the shortest distance between the corresponding reference object and the corresponding vibration measuring point is collected and marked as a reference distance detection value, and the reference distance detection value is numerically compared with the corresponding preset reference distance detection threshold. If the reference distance detection value does not exceed the corresponding preset reference distance detection threshold, it indicates that the distance between the corresponding vibration measuring point and the corresponding reference object is relatively close and is easily affected by it, and the corresponding vibration measuring point is marked as an unreasonable layout measuring point;

[0046] If there are unreasonable measurement points on the surface of the power transformer oil tank, a layout warning signal is generated; if there are no unreasonable measurement points on the surface of the power transformer oil tank, several detection areas are set on the surface of the power transformer oil tank, and analysis is performed to determine whether the corresponding detection area is a non-optimal layout area. If there is a non-optimal layout area on the surface of the power transformer oil tank, a layout warning signal is generated; the analysis and judgment method of the non-optimal layout area is as follows:

[0047] The number of vibration measuring points in the corresponding detection area is collected and marked as the measurement point number check value, and the measurement point number check value is compared with the corresponding preset measurement point number check value range. If the measurement point number check value is not within the corresponding preset measurement point number check value range, it indicates that the layout of the vibration measuring points in the corresponding detection area is unreasonable, and the corresponding detection area is marked as a non-optimal layout area;

[0048] If the number of measurement points is within the corresponding preset number of measurement points, the distance between any two groups of vibration measurement points in the corresponding detection area is marked as the point distance detection value, and the point distance detection value is numerically compared with the preset point distance detection threshold. If the point distance detection value does not exceed the preset point distance detection threshold, indicating that the distance between the corresponding two vibration measurement points is too small, the corresponding point distance detection value is marked as the point distance abnormal value.

[0049] The number of point distance anomaly values ​​in the corresponding detection area is obtained and the ratio thereof is calculated with the number of point distance detection values ​​to obtain the point distance anomaly value, and the difference between the point distance anomaly value and the preset point distance detection threshold is marked as the point distance amplitude detection value (only the numerical value is taken, regardless of positive or negative), the point distance amplitude condition value is obtained by averaging all the point distance anomaly values, and the point distance amplitude detection value with the largest numerical value is marked as the point distance amplitude analysis value;

[0050] The regional distribution evaluation value QF is obtained by numerically calculating the point distance abnormality value QW, the point distance amplitude value QS and the point distance amplitude analysis value QP through the formula QF=rg1*QW+(rg2*QS+rg3*QP) / 2; wherein rg1, rg2 and rg3 are preset proportional coefficients, rg1>rg2>rg3>0; and the larger the value of the regional distribution evaluation value QF is, the more unreasonable the distribution of vibration measuring points in the corresponding detection area is;

[0051] The regional distribution evaluation value QF is numerically compared with the corresponding preset regional distribution evaluation threshold. If the regional distribution evaluation value QF exceeds the preset regional distribution evaluation threshold, it indicates that the layout of the vibration measurement points in the corresponding detection area is unreasonable, and the corresponding detection area is marked as a non-optimal layout area.

[0052] Step 2: If no layout warning signal is generated, the vibration signal acquisition module uses a high-precision vibration sensor to synchronously acquire vibration signals at each vibration measurement point, and the acquired vibration signals are sent to the signal preprocessing module via the processor.

[0053] Step three: The signal preprocessing module preprocesses the collected vibration signal and sends the preprocessed vibration signal to the feature extraction and analysis module through the processor. The preprocessing operation includes filtering, denoising and normalization steps to improve the accuracy and reliability of signal analysis.

[0054] Step 4: The feature extraction and analysis module analyzes the pre-processed vibration signal and sends the analysis results to the sensitive area evaluation and selection module through the processor. For example, the vibration signal is subjected to spectrum analysis to extract the amplitude and phase information of key frequency components (such as 100Hz and its harmonics). At the same time, characteristic parameters such as the energy distribution and main frequency ratio of the vibration signal are calculated to fully reflect the characteristics of the vibration signal.

[0055] The vibration signal and the load current signal of the transformer are collected synchronously, and the correlation coefficient between the vibration signal and the load current is calculated, focusing on the vibration frequency components that are closely related to the load current changes. These components can reflect the stress condition and operating status of the winding.

[0056] Step 5. The sensitive area assessment selection module constructs a sensitive area assessment model, calculates a comprehensive sensitivity score for each vibration measurement point through the sensitive area assessment model and based on the analysis results of the feature extraction and analysis module, selects the location of the vibration measurement point with the highest score as the sensitive area according to the comprehensive sensitivity score result, and sends the selected sensitive area to the supervision terminal through the processor.

[0057] Specifically: According to the vibration characteristics evaluation results, combined with the correlation between the 100Hz component amplitude of the vibration signal and the effective value of the load current, the attenuation of vibration energy and the main component ratio of the vibration signal, the weight coefficient is determined and the comprehensive vibration evaluation index is calculated, that is, the comprehensive sensitivity score. Finally, the area with the highest comprehensive vibration evaluation index is selected as the sensitive area.

[0058] The present invention comprehensively considers various characteristic indicators of vibration signals, selects sensitive areas through scientific methods, avoids arranging measuring points in areas where vibration energy attenuation is serious or the ability to reflect winding vibration is poor, optimizes the monitoring layout, improves monitoring efficiency, ensures that the selected area can accurately reflect the vibration characteristics and status of the transformer winding, and improves the accuracy of monitoring.

[0059] Embodiment 2: Figure 3 As shown, the difference between this embodiment and the first embodiment is that the processor is communicatively connected to the sensitive area reselection warning module, and the sensitive area reselection warning module is used to re-analyze the selection urgency of the sensitive area of ​​the winding vibration on the surface of the power transformer oil tank, and to determine whether to generate a reselection warning signal through analysis;

[0060] The reselection warning signal is sent to the supervision terminal through the processor. When the supervision terminal receives the reselection warning signal, it issues a corresponding warning to remind the management personnel to reselect the sensitive area of ​​the winding vibration on the surface of the power transformer oil tank in time, so as to ensure the subsequent monitoring effect, reduce the difficulty of management, and have a high degree of intelligence. The specific analysis process of the sensitive area reselection warning module is as follows:

[0061] The selection time of the sensitive area adjacent to the last winding vibration on the surface of the power transformer oil tank is obtained and marked as the neighboring selection time, and the interval between the neighboring selection time and the current time is marked as the neighboring selection interval value, and the neighboring selection interval value is numerically compared with the preset neighboring selection interval threshold. If the neighboring selection interval value exceeds the preset neighboring selection interval threshold, it indicates that the interval time from the last selection of the sensitive area is long, and the sensitive area should be reselected, then a reselection warning signal is generated.

[0062] If the neighbor selection interval value does not exceed the preset neighbor selection interval threshold, the deformation values ​​of several detection points on the surface of the power transformer oil tank within the interval between the neighbor selection time and the current time are collected, and the deformation value of the corresponding detection point is numerically compared with the preset deformation threshold. If the deformation value exceeds the preset deformation threshold, it indicates that the deformation of the corresponding detection point is more serious, and the corresponding detection point is marked as a variation point;

[0063] The number of variation points is obtained and the ratio thereof is calculated with the total number of detection points to obtain the variation detection value, and the deformation values ​​of all detection points are averaged to obtain the deformation analysis value, and the deformation value with the largest value is marked as the deformation amplitude value;

[0064] The variation detection value XW, the deformation analysis value XL and the deformation amplitude value XP are numerically calculated by the formula XF=kp1*XW+(kp2*XL+kp3*XP) / 2 to obtain the adjacent shape measurement value XF; wherein, kp1, kp2 and kp3 are preset proportional coefficients, kp1>kp2>kp3>0; and the larger the value of the adjacent shape measurement value XF, the greater the shape change of the surface of the power transformer oil tank within the corresponding interval time;

[0065] The impact force of the collision impact process is collected through the pressure sensor, and the maximum impact force, the average impact force, and the duration of the collision impact are obtained and marked as the collision condition value, and the maximum impact force, the average impact force and the collision condition value are recorded;

[0066] By formula The collision condition value TN, the maximum impact force TK and the average impact force TL are numerically calculated to obtain the collision damage inspection value PX; wherein eg1, eg2, eg3 are preset proportional coefficients with values ​​greater than zero, and the larger the value of the collision damage inspection value PX, the greater the damage caused to the surface of the power transformer oil tank by the corresponding collision impact process;

[0067] The collision damage detection value PX is numerically compared with the preset collision damage detection threshold. If the collision damage detection value PX exceeds the preset collision damage detection threshold, it indicates that the corresponding collision impact process causes greater damage to the surface of the power transformer oil tank, and the corresponding collision impact process is marked as an abnormal collision process;

[0068] Obtain the total number of times the surface of the oil tank of the power transformer is impacted by objects during the interval between the adjacent selected time and the current time and mark it as the total frequency value of damage, and obtain the number of abnormal collision processes on the surface of the oil tank of the power transformer during the interval between the adjacent selected time and the current time and mark it as the total frequency value of abnormal collision;

[0069] The adjacent impact value LX is calculated by weighting and summing the damaged total frequency value LP and the abnormal collision total frequency value LS through the formula LX=(c1*LP+c2*LS) / 2; wherein c1 and c2 are preset weight coefficients, c2>c1>0; and the larger the value of the adjacent impact value LX, the greater the damage to the surface of the power transformer oil tank within the corresponding interval time;

[0070] The ratio of the neighbor selection interval value to the preset neighbor selection interval threshold is marked as the neighbor selection interval value; through the formula The neighboring partition occupancy value HY, the neighboring partition shape measurement value XF and the neighboring partition impact value LX are numerically calculated to obtain the reselection warning coefficient HX; wherein sq1, sq2, and sq3 are preset proportional coefficients with values ​​greater than zero, and the larger the value of the reselection warning coefficient HX, the more it is necessary to reselect the sensitive area in a timely manner;

[0071] The reselection warning coefficient HX is numerically compared with the preset reselection warning coefficient threshold. If the reselection warning coefficient HX exceeds the preset reselection warning coefficient threshold, it indicates that the sensitive area needs to be reselected in time, and a reselection warning signal is generated.

[0072] The working principle of the present invention is as follows: when in use, the positions of all vibration measuring points are acquired through the measuring point layout scanning module, and the layout rationality analysis is performed to determine whether to generate a layout warning signal. When the layout warning signal is generated, the management personnel are reminded to adjust the layout of the vibration measuring points, which is beneficial to ensure the accuracy of the finally selected sensitive area. When the layout warning signal is not generated, the vibration signal is synchronously collected at each vibration measuring point through the vibration signal acquisition module, and the collected vibration signal is preprocessed and feature extraction and analysis is performed. A comprehensive sensitivity score is calculated for each vibration measuring point through the sensitive area evaluation model and based on the analysis results of the feature extraction and analysis module, and the location of the vibration measuring point with the highest score is selected as the sensitive area to ensure that the selected area can accurately reflect the vibration characteristics and status of the transformer winding, thereby significantly improving the monitoring accuracy.

[0073] The above formulas are all dimensionless and numerical calculations. The formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formula are set by technicians in this field according to actual conditions. The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the contents of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that technicians in the relevant technical field can understand and use the present invention well. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A method for selecting sensitive areas of winding vibration on the surface of a power transformer oil tank, characterized in that: The following steps are involved: Step 1: The measurement point layout scanning module collects the surface image of the power transformer oil tank, obtains the positions of all vibration measurement points based on the surface image, and determines whether to generate a layout warning signal through layout rationality analysis; Step 2: If no layout warning signal is generated, a vibration signal is synchronously collected at each vibration measuring point using a high-precision vibration sensor through a vibration signal collection module, and the collected vibration signal is sent to a signal preprocessing module through a processor; Step 3: The signal preprocessing module preprocesses the collected vibration signal, and sends the preprocessed vibration signal to the feature extraction and analysis module through the processor; Step 4: The feature extraction and analysis module analyzes the preprocessed vibration signal and sends the analysis result to the sensitive area evaluation and selection module through the processor; Step 5: The sensitive area assessment selection module constructs a sensitive area assessment model, calculates a comprehensive sensitivity score for each vibration measurement point through the sensitive area assessment model and based on the analysis results of the feature extraction and analysis module, selects the location of the vibration measurement point with the highest score as the sensitive area, and sends the selected sensitive area to the supervision terminal through the processor; In step 1, the specific analysis method of layout rationality analysis is as follows: Based on the surface image of the power transformer oil tank, the high-impact structure and high-impact accessories on the oil tank are identified; the high-impact structure includes reinforcing ribs and support frames, and the high-impact accessories include oil inlets and outlets, cooler interfaces and valves; the high-impact structure and high-impact accessories are marked as reference objects, and the vibration measurement points distributed on the corresponding reference objects are marked as unreasonable layout measurement points; If there is no vibration measuring point on the corresponding reference object, the shortest distance between the corresponding reference object and the corresponding vibration measuring point is collected and marked as a reference distance detection value, and the reference distance detection value is numerically compared with the corresponding preset reference distance detection threshold. If the reference distance detection value does not exceed the corresponding preset reference distance detection threshold, the corresponding vibration measuring point is marked as an unreasonable layout measuring point; If there are unreasonable layout measurement points on the surface of the power transformer oil tank, a layout warning signal will be generated.

2. The method for selecting a sensitive area of ​​winding vibration on the surface of a power transformer oil tank according to claim 1, characterized in that: If there are no unreasonable layout measurement points on the surface of the power transformer oil tank, several detection areas are set on the surface of the power transformer oil tank. Through analysis, it is determined whether the corresponding detection area is a non-optimal layout area. If there is a non-optimal layout area on the surface of the power transformer oil tank, a layout warning signal is generated.

3. The method for selecting the sensitive area of ​​winding vibration on the surface of the power transformer oil tank according to claim 2 is characterized in that: The analysis and judgment methods for the layout of non-optimal areas are as follows: The number of vibration measurement points in the corresponding detection area is collected and marked as the measurement point number inspection value, and the measurement point number inspection value is compared with the corresponding preset measurement point number inspection value range. If the measurement point number inspection value is not within the corresponding preset measurement point number inspection value range, the corresponding detection area is marked as a layout non-optimal area.

4. The method for selecting the sensitive area of ​​winding vibration on the surface of the power transformer oil tank according to claim 3 is characterized in that: If the number of measuring points is within the corresponding preset number of measuring points, the distance between any two groups of vibration measuring points in the corresponding detection area is marked as the point distance detection value, and the point distance detection value is numerically compared with the preset point distance detection threshold. If the point distance detection value does not exceed the preset point distance detection threshold, the corresponding point distance detection value is marked as the point distance abnormal value. The number of point distance anomaly values ​​in the corresponding detection area is obtained and the ratio thereof is calculated with the number of point distance detection values ​​to obtain the point distance anomaly value, and the difference between the point distance anomaly value and the preset point distance detection threshold is marked as the point distance amplitude detection value, the mean of all point distance anomaly values ​​is calculated to obtain the point distance amplitude value, and the point distance amplitude detection value with the largest value is marked as the point distance amplitude analysis value; The regional distribution evaluation value is obtained by numerically calculating the point distance anomaly value, the point distance amplitude value and the point distance amplitude analysis value, and the regional distribution evaluation value is numerically compared with the corresponding preset regional distribution evaluation threshold. If the regional distribution evaluation value exceeds the preset regional distribution evaluation threshold, the corresponding detection area will be marked as a non-optimal layout area.

5. The method for selecting the sensitive area of ​​winding vibration on the surface of the power transformer oil tank according to claim 1, characterized in that: The processor is communicatively connected to the sensitive area reselection warning module. The sensitive area reselection warning module is used to re-analyze the selection urgency of the sensitive area of ​​the winding vibration on the surface of the power transformer oil tank, determine whether to generate a reselection warning signal through analysis, and send the reselection warning signal to the supervision terminal through the processor. When the supervision terminal receives the reselection warning signal, it will issue a corresponding warning.

6. A method for selecting a sensitive area of ​​winding vibration on the surface of a power transformer oil tank according to claim 5, characterized in that: The specific analysis process of the sensitive area reselection warning module includes: The selection moment of the last sensitive area where the winding on the surface of the power transformer oil tank vibrates is obtained and marked as the neighboring selection moment, and the interval between the neighboring selection moment and the current moment is marked as the neighboring selection interval value. The neighboring selection interval value is numerically compared with the preset neighboring selection interval threshold. If the neighboring selection interval value exceeds the preset neighboring selection interval threshold, a reselection warning signal is generated.

7. A method for selecting a sensitive area of ​​winding vibration on the surface of a power transformer oil tank according to claim 6, characterized in that: If the neighbor selection interval value does not exceed the preset neighbor selection interval threshold, the neighbor selection shape measurement value and the neighbor selection impact value are obtained through analysis, and the ratio of the neighbor selection interval value to the preset neighbor selection interval threshold is marked as the neighbor selection interval occupancy value; The re-election warning coefficient is obtained by numerically calculating the neighboring separation occupancy value, the neighboring separation shape measurement value and the neighboring separation impact value, and the re-election warning coefficient is numerically compared with the preset re-election warning coefficient threshold. If the re-election warning coefficient exceeds the preset re-election warning coefficient threshold, a re-election warning signal is generated.

8. The method for selecting the sensitive area of ​​winding vibration on the surface of the power transformer oil tank according to claim 7, characterized in that: The analysis and acquisition method of the apex shape measurement value is as follows: The deformation value of several detection points on the surface of the oil tank of the power transformer within the interval between the adjacent selected time and the current time is collected by the deformation sensor, and the deformation value of the corresponding detection point is numerically compared with the preset deformation threshold value. If the deformation value exceeds the preset deformation threshold value, the corresponding detection point is marked as a variation point; The number of variation points is obtained and the ratio is calculated with the total number of detection points to obtain the variation detection value, and the deformation values ​​of all detection points are averaged to obtain the deformation analysis value, and the deformation value with the largest value is marked as the deformation amplitude value. The adjacent shape measurement value is obtained by numerically calculating the variation detection value, deformation analysis value and deformation amplitude value.

9. A method for selecting a sensitive area of ​​winding vibration on the surface of a power transformer oil tank according to claim 8, characterized in that: The analysis and acquisition method of the adjacent impact value is as follows: When the surface of the oil tank of the power transformer is impacted by an object, the impact force of the collision impact process is collected through a pressure sensor, the maximum impact force, the average impact force, and the duration of the collision impact in the collision impact process are obtained and marked as the collision condition value, the maximum impact force, the average impact force and the collision condition value are numerically calculated to obtain the collision damage detection value, the collision damage detection value is numerically compared with the preset collision damage detection threshold, if the collision damage detection value exceeds the preset collision damage detection threshold, the corresponding collision impact process is marked as an abnormal collision process; The total number of times the surface of the power transformer oil tank is impacted by objects during the interval between the adjacent selected moment and the current moment is obtained and marked as the total frequency value of damage, and the number of abnormal collision processes on the surface of the power transformer oil tank during the interval between the adjacent selected moment and the current moment is obtained and marked as the total frequency value of abnormal collisions. The total frequency value of damage and the total frequency value of abnormal collisions are weighted and summed to obtain the adjacent impact value.

Images