Sound wave analysis method and system applied to equipment fault location

Abstract

The invention discloses a sound wave analysis method and system applied to equipment fault positioning, and relates to the technical field of power grid equipment fault analysis method.Conformance comparison is conducted on each normal frequency spectrum curve set and a basic frequency spectrum curve set, and the most conforming normal frequency spectrum curve set with the highest conforming degree is determined based on the comparison result; marking frequency spectrum sections corresponding to the difference factors in the basic frequency spectrum curve groups, determining an abnormal position block where an abnormality occurs on the equipment based on a time sequence relation of the frequency spectrum sections marked with the difference factors in different basic frequency spectrum curve groups, and analyzing the frequency spectrum sections marked with the abnormality on the basic frequency spectrum curve groups. According to the technical scheme of the invention, the method achieves the effect of positioning the abnormal position block of the equipment through sound waves, analyzes the abnormal features of the abnormal position block, and facilitates the timely making of corresponding maintenance measures for the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid equipment fault analysis, and in particular to an acoustic wave analysis method and system for locating equipment faults. Background Art

[0002] In the power grid system, the normal operation of equipment is crucial to ensure the stability and safety of power supply. However, equipment may fail due to long-term operation, overload, environmental factors, etc. Traditional fault detection methods mostly rely on manual inspections and historical data of equipment. These methods are often time-consuming and labor-intensive, and it is difficult to discover, locate, and analyze faults in real time. Summary of the invention

[0003] The purpose of the present invention is to provide a method and system capable of locating and analyzing faults of equipment in a power grid.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions: An acoustic wave analysis method for locating equipment faults, comprising: Based on the set position of the acoustic wave collection device relative to the equipment, a sound wave spatial collection array is constructed, the sound wave spatial collection array includes sound wave collection nodes at different spatial position nodes, and the sound wave data collected by the sound wave collection device are recorded separately in a manner of associating the sound wave collection nodes to obtain a plurality of sound wave data sets; The acoustic wave data set is analyzed by using a time-frequency analysis technique to obtain a number of basic spectrum curves, and a combination of basic spectrum curves belonging to the same acoustic wave data set is identified as a basic spectrum curve group; Several normal spectrum curve groups are preset for the normal operation state of the equipment, and each normal spectrum curve group is compared with the basic spectrum curve group for consistency, and based on the comparison results, the most consistent normal spectrum curve group with the highest degree of consistency is determined; Compare the most consistent normal spectrum curve group with the basic spectrum curve group to determine a number of difference factors. If the difference factor is within the preset normal range, it is determined that the device is normal. If the difference factor is not within the preset normal range, it is determined that the device is abnormal, and the spectrum segment corresponding to the difference factor is marked in the basic spectrum curve group; Based on the time sequence of the appearance of the spectrum segments marked with difference factors in different basic spectrum curve groups, the abnormal position block where the abnormality occurs on the equipment is determined, and the preset abnormal spectrum analysis model is used to analyze the spectrum segments marked with abnormalities on the basic spectrum curve group to determine the abnormal situation of the abnormal position block.

[0005] In some embodiments disclosed in the present invention, the method of comparing the consistency of each normal spectrum curve group with the basic spectrum curve group includes: Each normal spectrum curve in the normal spectrum curve group and each basic spectrum curve in the basic spectrum curve group are grouped and compared accordingly, and the basic spectrum curve is deformed within a preset range, and the normal spectrum curve and the basic spectrum curve in each spectrum curve comparison group are relatively dynamically translated on the time axis, and the normal spectrum curve and the basic spectrum curve in each spectrum curve comparison group are translated in the same manner, and after each dynamic translation, the overlap parameters of the normal spectrum curve and the basic spectrum curve are calculated in real time, and the translation mode corresponding to the highest overlap parameter is identified as the overlap translation mode; Based on the overlapping translation method, the normal spectrum curve and the basic spectrum curve are translated on the time axis, and based on the consistency characteristics between the translated normal spectrum curve and the basic spectrum curve, the sub-matching degree of the corresponding spectrum curve comparison group is determined, and based on the sub-matching degree of all spectrum curve comparison groups, the consistency degree of the normal spectrum curve group and the basic spectrum curve group is determined.

[0006] In an embodiment disclosed in the present invention, a method for determining the degree of consistency between a normal spectrum curve group and a basic spectrum curve group includes: Based on the equipment components preset corresponding to the normal spectrum curves in the spectrum curve comparison group, determining the matching weight coefficient of the spectrum curve comparison group, and combining the sub-matching degree of the spectrum curve comparison group, calculating the matching degree of the normal spectrum curve group and the basic spectrum curve group; The method for determining the sub-matching degree of the spectrum curve comparison group includes calculating the peak difference between the normal spectrum curve and the basic spectrum curve, and determining the number of spectrum line crossings between the normal spectrum curve and the basic spectrum curve within a preset time segment, and determining the sub-matching degree of the spectrum curve comparison group based on the peak difference and the number of spectrum line crossings; The expression for calculating the degree of conformity is: ; in, is the degree of correspondence between the spectrum curve comparison groups, is the corresponding weight coefficient of the i-th equipment component, that is, the corresponding weight coefficient of the i-th spectrum curve comparison group, is the sub-matching degree corresponding to the i-th spectrum curve comparison group; The expression for calculating the sub-matching degree is: ; in, is the first conformity adjustment coefficient, is the second conformity adjustment coefficient, is the peak difference amount affecting the judgment function, is the peak difference, are several preset peak difference intervals, According to the peak difference The peak difference interval , output a specific first degree of conformity parameter, is the influence judgment function of the number of spectrum line crossings, u is the number of spectrum line crossings, is a preset number of spectral line crossing intervals, According to the spectrum line crossing number u to which the spectrum line crossing number interval Output a specific second conformity parameter, b is a conformity adjustment constant.

[0007] In some embodiments disclosed in the present invention, the method for calculating the overlap parameter of the normal spectrum curve and the basic spectrum curve includes: Shift the time axis to the center axis of the normal spectrum curve and the basic spectrum curve; According to a preset time interval, a number of time mapping nodes are set on the time axis, and for each time mapping node, the corresponding normal spectrum curve value and the basic spectrum curve value are intercepted, and the normal spectrum curve value and the basic spectrum curve value belonging to the same time mapping node are recorded as a spectrum curve value group; Analyze each spectrum curve value group, if the normal spectrum curve value and the basic spectrum curve value are both positive or negative, then identify the smaller curve value between the normal spectrum curve value and the basic spectrum curve value as a sub-coincidence parameter, and the sub-coincidence parameter takes an absolute value; All sub-coincidence parameters are summed to obtain the coincidence parameters of the normal spectrum curve and the basic spectrum curve.

[0008] In some embodiments disclosed in the present invention, the method of determining a number of difference factors includes: The normal spectrum curves in the most consistent normal spectrum curve group and the basic spectrum curves in the basic spectrum curve group are matched one by one, and each pair of corresponding normal spectrum curves and basic spectrum curves is recorded as a spectrum curve comparison group; Deforming a basic spectrum curve in the basic spectrum curve in the spectrum curve comparison group, wherein the deformation method is to adjust the period and peak value of the basic spectrum curve within a preset range to obtain a deformed spectrum curve; Align the normal spectrum curve and the deformed spectrum curve, and construct a vertical scan line, gradually move the vertical scan line, and intercept the intersection of the vertical scan line and the normal spectrum curve and the deformed spectrum curve, record them as the normal spectrum curve value and the deformed spectrum curve value, and record the combination of the normal spectrum curve value and the deformed spectrum curve value as the analysis curve value group; According to the relationship before and after the vertical scanning line is scanned, the curve value group for analysis is sorted, and the curve value difference between the normal spectrum curve value and the deformed spectrum curve value in the curve value group for analysis is calculated. Based on the continuous performance characteristics of the curve value difference, the difference factor parameter of the basic spectrum curve relative to the normal spectrum curve is determined. If the difference factor parameter is greater than or equal to the preset value, the difference factor is marked on the segment corresponding to the basic spectrum curve.

[0009] In some embodiments disclosed in the present invention, based on the time sequence of occurrence of spectrum segments marked with difference factors in different basic spectrum curve groups, a method for determining an abnormal position block where an abnormality occurs on a device includes: A spatial coordinate system is established for the device, and based on the structural characteristics of the device, a corresponding virtual three-dimensional device structure is set in the spatial coordinate system, and the sound wave collection array is configured in the spatial coordinate system; Analyze the basic spectrum curve group corresponding to each device acquisition node respectively, determine the time node of each difference factor, and sort the difference factors according to the time relationship to obtain the difference factor sequence, and mark the time node of the difference factor; Compare the difference factor sequences corresponding to different equipment collection nodes, and calculate the occurrence time difference of the same difference factor between different equipment collection nodes; A number of component blocks are set inside the virtual three-dimensional device structure, and the sound wave conduction distance between each component block and different device acquisition nodes is gradually measured, and the difference between the sound wave conduction distances is calculated. If the difference between the sound wave conduction distances corresponding to a certain component block is consistent with the difference in occurrence time of the same difference factor, the component block is identified as an abnormal position block, and the difference factor is associated with the abnormal position block.

[0010] In some embodiments disclosed in the present invention, the method for constructing an abnormal spectrum graph analysis model includes: Establishing equipment abnormality test, including artificially setting abnormal faults for the corresponding components of the equipment based on the abnormal feature table set for each component, and using the acoustic wave collection device to collect abnormal acoustic wave data, and recording them separately according to the corresponding acoustic wave collection nodes to obtain several abnormal acoustic wave data sets; Analyze the abnormal sound wave data set to obtain several abnormal spectrum curves, intercept the spectrum segments on the abnormal spectrum curves, and associate the abnormal spectrum segments with abnormal features to obtain an abnormal feature-spectrum segment relationship group; The abnormal feature-spectrum segment relationship groups are analyzed to determine the feature label of each abnormal feature-spectrum segment relationship group.

[0011] In some embodiments disclosed in the present invention, the feature labels of the abnormal feature-spectrum segment relationship group include: the component to which the abnormal feature belongs, the average spectrum curve value of the spectrum segment, the upper and lower spectrum curve peaks of the spectrum segment, and the average spectrum curve period of the spectrum segment.

[0012] In some embodiments disclosed in the present invention, the method for classifying the spectrum segments marked with abnormalities on the basic spectrum curve group includes: Based on the determined abnormal position block, the abnormal feature-spectrum segment relationship group is screened for the first time, and based on the average spectrum curve value, upper and lower spectrum curve peaks and average spectrum curve period of the spectrum segment marked with abnormality on the basic spectrum curve group, the abnormal feature-spectrum segment relationship group is screened for the second time, and the screened abnormal feature-spectrum segment relationship group is identified as the reference abnormal feature-spectrum segment relationship group; The abnormal features in the reference abnormal feature-spectrum segment relationship group are identified as abnormal situations of the abnormal position block.

[0013] In some embodiments disclosed in the present invention, there is also disclosed an acoustic wave analysis system for locating equipment faults, including: The first module is used to construct an acoustic wave spatial acquisition array based on the set position of the acoustic wave acquisition device relative to the equipment, the acoustic wave spatial acquisition array includes acoustic wave acquisition nodes at different spatial position nodes, and the acoustic wave data collected by the acoustic wave acquisition device are recorded separately in a manner of associating the acoustic wave acquisition nodes to obtain a plurality of acoustic wave data sets; The second module is used to analyze the sound wave data set by using the time-frequency analysis technology to obtain a number of basic spectrum curves, and identify the combination of basic spectrum curves belonging to the same sound wave data set as a basic spectrum curve group; The third module is used to preset a number of normal spectrum curve groups for the normal operation state of the equipment, compare each normal spectrum curve group with the basic spectrum curve group for consistency, and determine the most consistent normal spectrum curve group with the highest degree of consistency based on the comparison results; The fourth module is used to compare the most consistent normal spectrum curve group with the basic spectrum curve group to determine a number of difference factors. If the difference factor is within the preset normal range, it is determined that the device is normal. If the difference factor is not within the preset normal range, it is determined that the device is abnormal, and the spectrum segment corresponding to the difference factor is marked in the basic spectrum curve group; The fifth module is used to determine the abnormal position block where the abnormality occurs on the equipment based on the time sequence of the appearance of the spectrum segments marked with difference factors in different basic spectrum curve groups, and use the preset abnormal spectrum analysis model to analyze the spectrum segments marked with abnormalities on the basic spectrum curve group to determine the abnormal situation of the abnormal position block.

[0014] The present invention discloses an acoustic wave analysis method and system for locating equipment faults, and relates to the technical field of power grid equipment fault analysis methods. Each normal spectrum curve group is respectively compared with a basic spectrum curve group for consistency, and based on the comparison result, the most consistent normal spectrum curve group with the highest degree of consistency is determined, and the spectrum segment corresponding to the difference factor is marked in the basic spectrum curve group. Based on the time sequence of the appearance of the spectrum segments marked with the difference factor in different basic spectrum curve groups, the abnormal position block where the abnormality occurs on the equipment is determined, and the spectrum segment marked with the abnormality on the basic spectrum curve group is analyzed to determine the abnormal situation of the abnormal position block. Through the above technical scheme, the present invention realizes the effect of locating the abnormal position block of the equipment by using acoustic waves, and analyzes the abnormal characteristics of the abnormal position block, which is conducive to timely taking corresponding maintenance measures for the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 The present invention proposes an acoustic wave analysis method and system for locating equipment faults. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0017] In order to achieve the above object, the present invention adopts the following technical solutions: See also Figure 1 , an acoustic wave analysis method for locating equipment faults, comprising: Step S100, based on the set position of the acoustic wave collection device relative to the equipment, construct an acoustic wave spatial collection array, the acoustic wave spatial collection array includes acoustic wave collection nodes at different spatial position nodes, and the acoustic wave data collected by the acoustic wave collection device are recorded separately in a manner of associating the acoustic wave collection nodes to obtain a plurality of acoustic wave data sets.

[0018] In this step, in order to fully capture the acoustic wave signals generated by the operation of the power grid equipment, acoustic wave collection nodes are installed at different spatial locations around the equipment to construct an acoustic wave spatial collection array; these nodes can collect acoustic wave data from multiple angles and distances, thereby obtaining more comprehensive acoustic wave information; the acoustic wave collection device will record these acoustic wave data and record them separately according to their association with each acoustic wave collection node to form multiple acoustic wave data sets; this recording method helps to accurately associate the acoustic wave data with specific location information during subsequent analysis.

[0019] Step S200: Analyze the sound wave data set using a time-frequency analysis technique to obtain a number of basic frequency spectrum curves, and identify a combination of basic frequency spectrum curves belonging to the same sound wave data set as a basic frequency spectrum curve group.

[0020] The acoustic wave data set contains the acoustic wave signals when the equipment is running, which carry rich information about the equipment status; time-frequency analysis technology can reveal the frequency content of the acoustic wave signals and their changes over time; by performing time-frequency analysis on each acoustic wave data set, a series of basic spectrum curves can be obtained, which show the frequency distribution characteristics of the acoustic wave signals; all basic spectrum curves belonging to the same acoustic wave data set are combined to form a basic spectrum curve group, which provides a data basis for subsequent fault diagnosis.

[0021] Acoustic signals are composed of waves of different frequencies, each of which may correspond to different operating states or fault modes of the equipment; the purpose of time-frequency analysis is to decompose complex acoustic signals into different frequency components and observe how these components change over time. Through time-frequency analysis, the spectrum information of acoustic signals can be obtained, which is crucial for fault diagnosis and location.

[0022] Step S300 , a plurality of normal spectrum curve groups are preset for the normal operation state of the device, each normal spectrum curve group is compared with the basic spectrum curve group for consistency, and based on the comparison result, the most consistent normal spectrum curve group with the highest degree of consistency is determined.

[0023] In order to determine whether the operating status of the equipment is normal, a set of standards is needed for reference. In this step, a series of normal spectrum curve groups are preset for the normal operating status of the equipment. These curve groups represent the acoustic characteristics of the equipment when it is working normally; then, we compare each normal spectrum curve group with the basic spectrum curve group for consistency, and through comparative analysis, find the normal spectrum curve group that best matches the actual collected data. Such a comparison helps to understand whether the current operating status of the equipment deviates from the normal range.

[0024] In some embodiments disclosed in the present invention, the method of comparing the consistency of each normal spectrum curve group with the basic spectrum curve group includes: Step S301, each normal spectrum curve in the normal spectrum curve group and each basic spectrum curve in the basic spectrum curve group are grouped and compared accordingly, and the basic spectrum curve is deformed within a preset range, and the normal spectrum curve and the basic spectrum curve in each spectrum curve comparison group are relatively dynamically translated on the time axis, and the normal spectrum curve and the basic spectrum curve in each spectrum curve comparison group are translated in the same manner, and the coincidence parameters of the normal spectrum curve and the basic spectrum curve after each dynamic translation are calculated in real time, and the translation mode corresponding to the highest coincidence parameter is identified as the coincidence translation mode.

[0025] In some embodiments disclosed in the present invention, the method for calculating the overlap parameter of the normal spectrum curve and the basic spectrum curve includes: Step S3011: translating the time axis to the position of the center axis of the normal spectrum curve and the basic spectrum curve.

[0026] Step S3012, according to the preset time interval, a number of time mapping nodes are set on the time axis, and for each time mapping node, the corresponding normal spectrum curve value and basic spectrum curve value are intercepted, and the normal spectrum curve value and basic spectrum curve value belonging to the same time mapping node are recorded as a spectrum curve value group.

[0027] Step S3013: Analyze each spectrum curve value group. If both the normal spectrum curve value and the basic spectrum curve value are positive or negative, the smaller curve value between the normal spectrum curve value and the basic spectrum curve value is identified as a sub-coincidence parameter, and the sub-coincidence parameter takes an absolute value.

[0028] Step S3014: sum all the sub-coincidence parameters to obtain the coincidence parameters of the normal spectrum curve and the basic spectrum curve.

[0029] Step S302, based on the overlapping translation method, the normal spectrum curve and the basic spectrum curve are translated on the time axis, and based on the consistency characteristics between the translated normal spectrum curve and the basic spectrum curve, the sub-conformity degree of the corresponding spectrum curve comparison group is determined, and based on the sub-conformity degree of all spectrum curve comparison groups, the conformity degree of the normal spectrum curve group and the basic spectrum curve group is determined.

[0030] In an embodiment disclosed in the present invention, a method for determining the degree of consistency between a normal spectrum curve group and a basic spectrum curve group includes: Step S3021, based on the equipment components preset corresponding to the normal spectrum curves in the spectrum curve comparison group, determine the matching weight coefficient of the spectrum curve comparison group, and calculate the matching degree of the normal spectrum curve group and the basic spectrum curve group in combination with the sub-matching degree of the spectrum curve comparison group.

[0031] Among them, the method for determining the sub-matching degree of the spectrum curve comparison group includes calculating the peak difference between the normal spectrum curve and the basic spectrum curve, and determining the number of spectrum line intersections between the normal spectrum curve and the basic spectrum curve within a preset time segment, and determining the sub-matching degree of the spectrum curve comparison group based on the peak difference and the number of spectrum line intersections.

[0032] The expression for calculating the degree of conformity is: .

[0033] in, is the degree of correspondence between the spectrum curve comparison groups, is the corresponding weight coefficient of the i-th equipment component, that is, the corresponding weight coefficient of the i-th spectrum curve comparison group, is the sub-matching degree corresponding to the i-th spectrum curve comparison group; The expression for calculating the sub-matching degree is: .

[0034] in, is the first conformity adjustment coefficient, is the second conformity adjustment coefficient, is the peak difference amount affecting the judgment function, is the peak difference, are several preset peak difference intervals, According to the peak difference The peak difference interval , output a specific first degree of conformity parameter, is the influence judgment function of the number of spectrum line crossings, u is the number of spectrum line crossings, is a preset number of spectral line crossing intervals, According to the spectrum line crossing number u to which the spectrum line crossing number interval Output a specific second conformity parameter, b is a conformity adjustment constant.

[0035] Step S400, compare the most consistent normal spectrum curve group with the basic spectrum curve group to determine a number of difference factors. If the difference factor is within the preset normal range, it is determined that the device is normal. If the difference factor is not within the preset normal range, it is determined that the device is abnormal, and the spectrum segment corresponding to the difference factor is marked in the basic spectrum curve group.

[0036] By comparing the most consistent normal spectrum curve group and the basic spectrum curve group, the difference factors can be determined, which reflect the difference between the actual data and the normal data. If the difference factor is within the preset normal range, it can be considered that the operation status of the equipment is normal. However, if the difference factor exceeds the normal range, this indicates that the equipment may be abnormal. In this case, the spectrum segments corresponding to the difference factor will be marked in the basic spectrum curve group, and these segments are where the potential fault signals are located.

[0037] In some embodiments disclosed in the present invention, the method of determining a number of difference factors includes: Step S401 : A one-to-one correspondence is made between the normal spectrum curves in the most consistent normal spectrum curve group and the basic spectrum curves in the basic spectrum curve group, and each pair of corresponding normal spectrum curves and basic spectrum curves is recorded as a spectrum curve comparison group.

[0038] Step S402 : deforming a basic spectrum curve in the basic spectrum curve in the spectrum curve comparison group, wherein the deformation method is to adjust the period and peak value of the basic spectrum curve within a preset range to obtain a deformed spectrum curve.

[0039] Step S403, align the normal spectrum curve and the deformed spectrum curve, and construct a vertical scan line, gradually move the vertical scan line, and intercept the intersection of the vertical scan line and the normal spectrum curve and the deformed spectrum curve, record them as the normal spectrum curve value and the deformed spectrum curve value, and record the combination of the normal spectrum curve value and the deformed spectrum curve value as the curve value group for analysis.

[0040] Step S404, sorting the analysis curve value group according to the relationship before and after the vertical scanning line is scanned, and calculating the curve value difference between the normal spectrum curve value and the deformed spectrum curve value in the analysis curve value group, and determining the difference factor parameter of the basic spectrum curve relative to the normal spectrum curve based on the continuous performance characteristics of the curve value difference, if the difference factor parameter is greater than or equal to the preset value, marking the difference factor for the segment corresponding to the basic spectrum curve.

[0041] Step S500, based on the time sequence of the appearance of the spectrum segments marked with difference factors in different basic spectrum curve groups, determine the abnormal position block where the abnormality occurs on the device, and use the preset abnormal spectrum analysis model to analyze the spectrum segments marked with abnormalities on the basic spectrum curve group to determine the abnormal situation of the abnormal position block.

[0042] The last step is to determine the specific location block where the abnormality occurs on the equipment; based on the time sequence of the spectrum segments marked with difference factors in different basic spectrum curve groups, the approximate location of the abnormality can be inferred. Then, using the preset abnormal spectrum analysis model, the spectrum segments marked with abnormalities are deeply analyzed to determine the specific abnormality of the abnormal location block; such analysis helps to discover faults, understand the nature and severity of the faults, and provide guidance for repair and maintenance.

[0043] In some embodiments disclosed in the present invention, based on the time sequence of occurrence of spectrum segments marked with difference factors in different basic spectrum curve groups, a method for determining an abnormal position block where an abnormality occurs on a device includes: Step S501, a space coordinate system is established for the device, and based on the structural characteristics of the device, a corresponding virtual three-dimensional device structure is set in the space coordinate system, and the sound wave collection array is configured in the space coordinate system.

[0044] Step S502, analyze the basic spectrum curve group corresponding to each device acquisition node respectively, determine the time node of each difference factor, sort the difference factors according to the time relationship, obtain a difference factor sequence, and mark the occurrence time node of the difference factor.

[0045] Step S503: compare the difference factor sequences corresponding to different equipment collection nodes, and calculate the occurrence time difference of the same difference factor between different equipment collection nodes.

[0046] Step S504, setting a number of component blocks within the virtual three-dimensional device structure, and gradually measuring the sound wave conduction distance between each component block and different device acquisition nodes, and calculating the difference between the sound wave conduction distances. If the difference between the sound wave conduction distances corresponding to a certain component block is consistent with the occurrence time difference of the same difference factor, then the component block is identified as an abnormal position block, and the difference factor is associated with the abnormal position block.

[0047] In some embodiments disclosed in the present invention, the method for constructing an abnormal spectrum graph analysis model includes: Step S505, establishing an equipment abnormality test, including artificially setting abnormal faults for the components corresponding to the equipment based on the abnormal feature table set for each component, and using the sound wave collection device to collect abnormal sound wave data, and recording them separately according to the corresponding sound wave collection nodes to obtain several abnormal sound wave data sets.

[0048] In this step, abnormal faults of equipment components are artificially set to simulate possible fault conditions that may occur in the actual operation of the equipment. These abnormal faults can be preset or determined based on the historical fault data of the equipment. After the abnormal faults are set, the acoustic wave data under these abnormal conditions are collected using an acoustic wave collection device. These data reflect the acoustic wave characteristics of equipment components under abnormal conditions and are very important for subsequent fault diagnosis and positioning.

[0049] Step S506 , analyzing the abnormal sound wave data set to obtain a number of abnormal spectrum curves, intercepting spectrum segments on the abnormal spectrum curves, and associating the abnormal spectrum segments with abnormal features to obtain an abnormal feature-spectrum segment relationship group.

[0050] After obtaining the sound wave data under abnormal conditions, it is necessary to perform time-frequency analysis on these data to extract abnormal spectrum curves; these spectrum curves show the frequency distribution of sound wave signals under abnormal conditions; by analyzing these spectrum curves, the spectrum segments related to the abnormal features can be intercepted; these abnormal spectrum segments are associated with the corresponding abnormal features to form abnormal feature-spectrum segment relationship groups; these relationship groups contain the sound wave feature information under abnormal conditions, which is crucial for constructing abnormal spectrum graph analysis models.

[0051] Step S507 , analyzing the abnormal feature-spectrum segment relationship groups, and determining a feature label of each abnormal feature-spectrum segment relationship group.

[0052] The last step is to analyze the abnormal feature-spectrum segment relationship group to determine the feature label of each relationship group; the feature label is a description of the abnormal spectrum segment, which contains information such as the component to which the abnormal feature belongs, the average spectrum curve value of the spectrum segment, the upper and lower spectrum curve peaks of the spectrum segment, and the average spectrum curve period of the spectrum segment; these feature labels help the model identify and interpret abnormal patterns in acoustic wave data, thereby accurately diagnosing and locating equipment faults.

[0053] In some embodiments disclosed in the present invention, the feature labels of the abnormal feature-spectrum segment relationship group include: the component to which the abnormal feature belongs, the average spectrum curve value of the spectrum segment, the upper and lower spectrum curve peaks of the spectrum segment, and the average spectrum curve period of the spectrum segment.

[0054] In some embodiments disclosed in the present invention, the method for classifying the spectrum segments marked with abnormalities on the basic spectrum curve group includes: Step S508, based on the determined abnormal position block, the abnormal feature-spectrum segment relationship group is screened for the first time, and based on the average spectrum curve value, upper and lower spectrum curve peaks and average spectrum curve period of the spectrum segment marked with abnormalities on the basic spectrum curve group, the abnormal feature-spectrum segment relationship group is screened for the second time, and the screened abnormal feature-spectrum segment relationship group is identified as the reference abnormal feature-spectrum segment relationship group.

[0055] Step S509: identifying the abnormal feature in the reference abnormal feature-spectrum segment relationship group as an abnormal situation of the abnormal position block.

[0056] In some embodiments disclosed in the present invention, there is also disclosed an acoustic wave analysis system for locating equipment faults, including: The first module is used to construct an acoustic wave spatial acquisition array based on the set position of the acoustic wave acquisition device relative to the equipment, the acoustic wave spatial acquisition array includes acoustic wave acquisition nodes at different spatial position nodes, and the acoustic wave data collected by the acoustic wave acquisition device are recorded separately in a manner of associating the acoustic wave acquisition nodes to obtain a plurality of acoustic wave data sets; The second module is used to analyze the sound wave data set by using the time-frequency analysis technology to obtain a number of basic spectrum curves, and identify the combination of basic spectrum curves belonging to the same sound wave data set as a basic spectrum curve group; The third module is used to preset a number of normal spectrum curve groups for the normal operation state of the equipment, compare each normal spectrum curve group with the basic spectrum curve group for consistency, and determine the most consistent normal spectrum curve group with the highest degree of consistency based on the comparison results; The fourth module is used to compare the most consistent normal spectrum curve group with the basic spectrum curve group to determine a number of difference factors. If the difference factor is within the preset normal range, it is determined that the device is normal. If the difference factor is not within the preset normal range, it is determined that the device is abnormal, and the spectrum segment corresponding to the difference factor is marked in the basic spectrum curve group; The fifth module is used to determine the abnormal position block where the abnormality occurs on the equipment based on the time sequence of the appearance of the spectrum segments marked with difference factors in different basic spectrum curve groups, and use the preset abnormal spectrum analysis model to analyze the spectrum segments marked with abnormalities on the basic spectrum curve group to determine the abnormal situation of the abnormal position block.

[0057] The present invention discloses an acoustic wave analysis method and system for locating equipment faults, and relates to the technical field of power grid equipment fault analysis methods. Each normal spectrum curve group is respectively compared with a basic spectrum curve group for consistency, and based on the comparison result, the most consistent normal spectrum curve group with the highest degree of consistency is determined, and the spectrum segment corresponding to the difference factor is marked in the basic spectrum curve group. Based on the time sequence of the appearance of the spectrum segments marked with the difference factor in different basic spectrum curve groups, the abnormal position block where the abnormality occurs on the equipment is determined, and the spectrum segment marked with the abnormality on the basic spectrum curve group is analyzed to determine the abnormal situation of the abnormal position block. Through the above technical scheme, the present invention realizes the effect of locating the abnormal position block of the equipment by using acoustic waves, and analyzes the abnormal characteristics of the abnormal position block, which is conducive to timely taking corresponding maintenance measures for the equipment.

[0058] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An acoustic wave analysis method for locating equipment faults, characterized in that: include: Based on the set position of the acoustic wave collection device relative to the equipment, a sound wave spatial collection array is constructed, the sound wave spatial collection array includes sound wave collection nodes at different spatial position nodes, and the sound wave data collected by the sound wave collection device are recorded separately in a manner of associating the sound wave collection nodes to obtain a plurality of sound wave data sets; The acoustic wave data set is analyzed by using a time-frequency analysis technique to obtain a number of basic spectrum curves, and a combination of basic spectrum curves belonging to the same acoustic wave data set is identified as a basic spectrum curve group; Several normal spectrum curve groups are preset for the normal operation state of the equipment, and each normal spectrum curve group is compared with the basic spectrum curve group for consistency, and based on the comparison results, the most consistent normal spectrum curve group with the highest degree of consistency is determined; Compare the most consistent normal spectrum curve group with the basic spectrum curve group to determine a number of difference factors. If the difference factor is within the preset normal range, it is determined that the device is normal. If the difference factor is not within the preset normal range, it is determined that the device is abnormal, and the spectrum segment corresponding to the difference factor is marked in the basic spectrum curve group; Based on the time sequence of the appearance of the spectrum segments marked with difference factors in different basic spectrum curve groups, the abnormal position block where the abnormality occurs on the equipment is determined, and the preset abnormal spectrum analysis model is used to analyze the spectrum segments marked with abnormalities on the basic spectrum curve group to determine the abnormal situation of the abnormal position block.

2. The acoustic wave analysis method for locating equipment faults according to claim 1, characterized in that: The method of comparing the consistency of each normal spectrum curve group with the basic spectrum curve group includes: Each normal spectrum curve in the normal spectrum curve group and each basic spectrum curve in the basic spectrum curve group are grouped and compared accordingly, and the basic spectrum curve is deformed within a preset range, and the normal spectrum curve and the basic spectrum curve in each spectrum curve comparison group are relatively dynamically translated on the time axis, and the normal spectrum curve and the basic spectrum curve in each spectrum curve comparison group are translated in the same manner, and after each dynamic translation, the overlap parameters of the normal spectrum curve and the basic spectrum curve are calculated in real time, and the translation mode corresponding to the highest overlap parameter is identified as the overlap translation mode; Based on the overlapping translation method, the normal spectrum curve and the basic spectrum curve are translated on the time axis, and based on the consistency characteristics between the translated normal spectrum curve and the basic spectrum curve, the sub-conformity degree of the corresponding spectrum curve comparison group is determined, and based on the sub-conformity degree of all spectrum curve comparison groups, the consistency degree of the normal spectrum curve group and the basic spectrum curve group is determined.

3. The acoustic wave analysis method for locating equipment faults according to claim 2, characterized in that: Methods for determining the degree of conformity between the normal spectrum curve group and the basic spectrum curve group include: Based on the equipment components preset corresponding to the normal spectrum curves in the spectrum curve comparison group, determining the matching weight coefficient of the spectrum curve comparison group, and combining the sub-matching degree of the spectrum curve comparison group, calculating the matching degree of the normal spectrum curve group and the basic spectrum curve group; The method for determining the sub-matching degree of the spectrum curve comparison group includes calculating the peak difference between the normal spectrum curve and the basic spectrum curve, and determining the number of spectrum line crossings between the normal spectrum curve and the basic spectrum curve within a preset time segment, and determining the sub-matching degree of the spectrum curve comparison group based on the peak difference and the number of spectrum line crossings; The expression for calculating the degree of conformity is: ； in, is the degree of correspondence between the spectrum curve comparison groups, is the corresponding weight coefficient of the i-th equipment component, that is, the corresponding weight coefficient of the i-th spectrum curve comparison group, is the sub-matching degree corresponding to the i-th spectrum curve comparison group; The expression for calculating the sub-matching degree is: ； in, is the first conformity adjustment coefficient, is the second conformity adjustment coefficient, is the peak difference amount affecting the judgment function, is the peak difference, are several preset peak difference intervals, According to the peak difference The peak difference interval , output a specific first degree of conformity parameter, is the influence judgment function of the number of spectrum line crossings, u is the number of spectrum line crossings, is a preset number of spectral line crossing intervals, According to the spectrum line crossing number u to which the spectrum line crossing number interval Output a specific second conformity parameter, b is a conformity adjustment constant.

4. The acoustic wave analysis method for locating equipment faults according to claim 2, characterized in that: The method for calculating the coincidence parameter of the normal spectrum curve and the basic spectrum curve includes: Shift the time axis to the center axis of the normal spectrum curve and the basic spectrum curve; According to a preset time interval, a number of time mapping nodes are set on the time axis, and for each time mapping node, the corresponding normal spectrum curve value and the basic spectrum curve value are intercepted, and the normal spectrum curve value and the basic spectrum curve value belonging to the same time mapping node are recorded as a spectrum curve value group; Analyze each spectrum curve value group, if the normal spectrum curve value and the basic spectrum curve value are both positive or negative, then identify the smaller curve value between the normal spectrum curve value and the basic spectrum curve value as a sub-coincidence parameter, and the sub-coincidence parameter takes an absolute value; All sub-coincidence parameters are summed to obtain the coincidence parameters of the normal spectrum curve and the basic spectrum curve.

5. The acoustic wave analysis method for locating equipment faults according to claim 1, characterized in that: Methods for determining several differentiating factors include: The normal spectrum curves in the most consistent normal spectrum curve group and the basic spectrum curves in the basic spectrum curve group are matched one by one, and each pair of corresponding normal spectrum curves and basic spectrum curves is recorded as a spectrum curve comparison group; Deforming a basic spectrum curve in the basic spectrum curve in the spectrum curve comparison group, wherein the deformation method is to adjust the period and peak value of the basic spectrum curve within a preset range to obtain a deformed spectrum curve; Align the normal spectrum curve and the deformed spectrum curve, and construct a vertical scan line, gradually move the vertical scan line, and intercept the intersection of the vertical scan line and the normal spectrum curve and the deformed spectrum curve, record them as the normal spectrum curve value and the deformed spectrum curve value, and record the combination of the normal spectrum curve value and the deformed spectrum curve value as the analysis curve value group; According to the relationship before and after the vertical scanning line is scanned, the curve value group for analysis is sorted, and the curve value difference between the normal spectrum curve value and the deformed spectrum curve value in the curve value group for analysis is calculated. Based on the continuous performance characteristics of the curve value difference, the difference factor parameter of the basic spectrum curve relative to the normal spectrum curve is determined. If the difference factor parameter is greater than or equal to the preset value, the difference factor is marked on the segment corresponding to the basic spectrum curve.

6. The acoustic wave analysis method for locating equipment faults according to claim 1, characterized in that: The method for determining the abnormal position block where the abnormality occurs on the device based on the time sequence relationship of the frequency spectrum segments marked with the difference factors in different basic frequency spectrum curve groups includes: A spatial coordinate system is established for the device, and based on the structural characteristics of the device, a corresponding virtual three-dimensional device structure is set in the spatial coordinate system, and the sound wave collection array is configured in the spatial coordinate system; Analyze the basic spectrum curve group corresponding to each device acquisition node respectively, determine the time node of each difference factor, and sort the difference factors according to the time relationship to obtain the difference factor sequence, and mark the time node of the difference factor; Compare the difference factor sequences corresponding to different equipment collection nodes, and calculate the occurrence time difference of the same difference factor between different equipment collection nodes; A number of component blocks are set inside the virtual three-dimensional device structure, and the sound wave conduction distance between each component block and different device acquisition nodes is gradually measured, and the difference between the sound wave conduction distances is calculated. If the difference between the sound wave conduction distances corresponding to a certain component block is consistent with the difference in occurrence time of the same difference factor, the component block is identified as an abnormal position block, and the difference factor is associated with the abnormal position block.

7. The acoustic wave analysis method for equipment fault location according to claim 1, characterized in that: Methods for constructing an abnormal spectrum analysis model include: Establishing equipment abnormality test, including artificially setting abnormal faults for the corresponding components of the equipment based on the abnormal feature table set for each component, and using the acoustic wave collection device to collect abnormal acoustic wave data, and recording them separately according to the corresponding acoustic wave collection nodes to obtain several abnormal acoustic wave data sets; Analyze the abnormal sound wave data set to obtain several abnormal spectrum curves, intercept the spectrum segments on the abnormal spectrum curves, and associate the abnormal spectrum segments with abnormal features to obtain an abnormal feature-spectrum segment relationship group; The abnormal feature-spectrum segment relationship groups are analyzed to determine the feature label of each abnormal feature-spectrum segment relationship group.

8. The acoustic wave analysis method for locating equipment faults according to claim 7, characterized in that: The feature labels of the abnormal feature-spectrum segment relationship group include: the component to which the abnormal feature belongs, the average spectrum curve value of the spectrum segment, the upper and lower spectrum curve peaks of the spectrum segment, and the average spectrum curve period of the spectrum segment.

9. The acoustic wave analysis method for locating equipment faults according to claim 7, characterized in that: The method for classifying the spectrum segments marked with abnormalities on the basic spectrum curve group includes: Based on the determined abnormal position block, the abnormal feature-spectrum segment relationship group is screened for the first time, and based on the average spectrum curve value, upper and lower spectrum curve peaks and average spectrum curve period of the spectrum segment marked with abnormality on the basic spectrum curve group, the abnormal feature-spectrum segment relationship group is screened for the second time, and the screened abnormal feature-spectrum segment relationship group is identified as the reference abnormal feature-spectrum segment relationship group; The abnormal features in the reference abnormal feature-spectrum segment relationship group are identified as abnormal situations of the abnormal position block.

10. An acoustic wave analysis system for locating equipment faults, characterized in that: include: The first module is used to construct an acoustic wave spatial acquisition array based on the set position of the acoustic wave acquisition device relative to the equipment, the acoustic wave spatial acquisition array includes acoustic wave acquisition nodes at different spatial position nodes, and the acoustic wave data collected by the acoustic wave acquisition device are recorded separately in a manner of associating the acoustic wave acquisition nodes to obtain a plurality of acoustic wave data sets; The second module is used to analyze the sound wave data set by using the time-frequency analysis technology to obtain a number of basic spectrum curves, and identify the combination of basic spectrum curves belonging to the same sound wave data set as a basic spectrum curve group; The third module is used to preset a number of normal spectrum curve groups for the normal operation state of the equipment, compare each normal spectrum curve group with the basic spectrum curve group for consistency, and determine the most consistent normal spectrum curve group with the highest degree of consistency based on the comparison results; The fourth module is used to compare the most consistent normal spectrum curve group with the basic spectrum curve group to determine a number of difference factors. If the difference factor is within the preset normal range, it is determined that the device is normal. If the difference factor is not within the preset normal range, it is determined that the device is abnormal, and the spectrum segment corresponding to the difference factor is marked in the basic spectrum curve group; The fifth module is used to determine the abnormal position block where the abnormality occurs on the equipment based on the time sequence of the appearance of the spectrum segments marked with difference factors in different basic spectrum curve groups, and use the preset abnormal spectrum analysis model to analyze the spectrum segments marked with abnormalities on the basic spectrum curve group to determine the abnormal situation of the abnormal position block.

Images