Method for evaluating running state of electrical equipment based on joint detection of sound and vibration

Abstract

This invention discloses a method for evaluating the operating status of electrical equipment based on joint sound and vibration detection, belonging to the field of power monitoring and diagnostic technology. The method includes inputting a set of abnormal sound and vibration events into a competitive resolution model for the sources of these anomalies. Based on the source types—namely, abnormalities in the target equipment itself, external sound source interference, external vibration impact, structural propagation anomalies, and acquisition anomalies—the method performs source matching and competitive screening of abnormal sound segments, abnormal vibration segments, and their corresponding operating conditions, generating consistent abnormal source data for sound and vibration operating conditions. This invention corrects for anomaly sources using sound and vibration energy conservation deviation data, classifying external sound source interference, external vibration impact, structural propagation anomalies, and acquisition anomalies into non-equipment anomaly records, and classifying equipment anomaly sources into equipment anomaly evaluation records. This makes the operating status evaluation results more accurately reflect changes in the target electrical equipment's inherent anomalies, improving the reliability of early warning.

Description

Technical Field

[0001] This invention relates to the field of power monitoring and diagnostic technology, and in particular to a method for evaluating the operating status of electrical equipment based on the combined detection of sound and vibration. Background Technology

[0002] With the development of online monitoring and intelligent operation and maintenance methods for power equipment, sound detection and vibration detection are gradually being used to identify the operating status of motors, switchgear, circuit breakers, transformers and contactors. Sound signals can reflect acoustic anomalies, impact sounds and friction sounds, while vibration signals can reflect structural response, mechanical impact and changes in operational stability. Combined sound and vibration detection can provide a data foundation for equipment condition evaluation and early warning of faults.

[0003] Existing methods for joint acoustic and vibration detection mainly focus on acquisition, filtering, frequency domain feature extraction, and fault type matching. They often directly use acoustic and vibration anomalies as the basis for equipment anomalies. In complex operating environments, external sound source interference, external vibration impact, structural propagation anomalies, and acquisition anomalies can all cause abnormal segments. If there is a lack of competitive screening of anomaly sources and verification of energy correspondence, the evaluation results may easily deviate from the actual state of the equipment itself. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a method for evaluating the operating status of electrical equipment based on joint detection of sound and vibration, which solves the problem of inaccurate evaluation of the target equipment's status caused by the mixed sources of sound and vibration anomalies under complex operating conditions.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides a method for evaluating the operating status of electrical equipment based on joint sound and vibration detection. The method includes: acquiring sound and vibration condition data and anomaly energy conversion benchmarks of the target electrical equipment; performing time alignment, condition segmentation, and filtering out invalid acquisition segments on the sound and vibration condition data to form synchronous sound and vibration segments; extracting abnormal sound segments and abnormal vibration segments based on the synchronous sound and vibration segments, and associating and organizing them according to the time of occurrence of the anomaly, the duration of the anomaly, and the corresponding relationship of the operating conditions to form a set of abnormal sound and vibration events; inputting the set of abnormal sound and vibration events into a competition-resolving model for the source of abnormal sound and vibration, and performing analysis on the abnormal sound segments, abnormal vibration segments, and the corresponding relationship of the operating conditions according to the source types of the target equipment itself (anomaly, external sound source interference, external vibration impact, structural propagation anomaly, and acquisition anomaly). Source matching and competitive screening generate consistent anomalous source data for acoustic and vibration operating conditions. Based on this consistent anomalous source data, the sound energy change and vibration energy change corresponding to each consistent anomalous source are extracted, and the correspondence between the sound energy change and vibration energy change is compared with the body's anomalous energy conversion benchmark to form acoustic and vibration energy conservation deviation data. Based on this deviation data, the consistent anomalous source data for acoustic and vibration operating conditions is corrected, and external sound source interference, external vibration impact, structural propagation anomalies, and acquisition anomalies are classified into non-body anomaly records. Anomaly sources that meet the body's anomalous energy conversion benchmark are classified into body anomaly evaluation records. Based on the non-body anomaly records and body anomaly evaluation records, the target electrical equipment operating status evaluation results are generated.

[0007] As a preferred embodiment of the electrical equipment operation status evaluation method based on joint sound and vibration detection described in this invention, the specific steps for forming the sound and vibration synchronization segment are as follows: The sound and vibration data of the acoustic and vibration conditions are extracted, and time alignment and condition segmentation are performed to obtain segmented acoustic and vibration conditions data. By using segmented acoustic and vibration data, invalid acquisition segments with sampling gaps, time misalignments, and acquisition interruptions are screened out. The retained sound segments, vibration segments, operating condition segments, and the body's abnormal energy conversion benchmark are correlated and recombined to form acoustic and vibration synchronization segments.

[0008] As a preferred embodiment of the electrical equipment operating status evaluation method based on joint sound and vibration detection according to the present invention, the specific steps for forming a set of abnormal sound and vibration events are as follows: Extract the sound data segment and vibration data segment of the sound-vibration synchronization segment, and perform fluctuation interval identification and abnormal start and end location respectively to obtain the sound abnormal segment, vibration abnormal segment and the corresponding abnormal occurrence time; Based on the abnormal sound segments, abnormal vibration segments and their corresponding occurrence times, the abnormal duration intervals and the corresponding working conditions are matched and organized to obtain the sound and vibration abnormality associated segments. By using the correlation segments of acoustic and vibration anomalies, corresponding acoustic and vibration anomaly segments under the same working condition are recombined to form a set of acoustic and vibration anomaly events.

[0009] As a preferred embodiment of the electrical equipment operating status evaluation method based on joint sound and vibration detection described in this invention, the specific steps for generating consistent abnormal source data of sound and vibration operating conditions are as follows: Based on the set of acoustic and vibration abnormal events, the acoustic and vibration abnormal source competition resolution model is used to extract the corresponding sound abnormal segment, vibration abnormal segment, abnormal occurrence time, abnormal duration interval and working condition correspondence for each abnormal event. Sound abnormal segments and vibration abnormal segments that belong to the same abnormal occurrence time, the same abnormal duration interval and the same working condition correspondence are paired and organized to obtain acoustic and vibration abnormal pairing data. Based on the sound and vibration anomaly pairing data, focusing on the source types of the target equipment body anomaly, external sound source interference, external vibration impact, structural propagation anomaly, and acquisition anomaly, source pointing mark and correspondence verification are performed on each set of sound and vibration anomaly pairing data to obtain candidate source matching data; Using candidate source matching data, the correspondence between sound abnormal segments, vibration abnormal segments, and operating conditions of each candidate source is collaboratively determined. Candidate sources with matching relationships are determined as consistent sound and vibration operating conditions and retained, while candidate sources with unmatched relationships are determined as inconsistent sound and vibration operating conditions and eliminated, thus generating consistent sound and vibration operating condition abnormal source data.

[0010] As a preferred embodiment of the electrical equipment operating status evaluation method based on joint sound and vibration detection described in this invention, the specific construction process of the sound and vibration anomaly source competition resolution model is as follows: A model for resolving competition among acoustic and vibration anomaly sources is constructed based on an acoustic and vibration anomaly pairing identification layer, an anomaly source competition discrimination layer, and a working condition consistency screening layer. The sound and vibration anomaly pairing and identification layer, based on the set of sound and vibration anomaly events, organizes the features of sound anomaly segments, vibration anomaly segments, anomaly occurrence time, anomaly duration interval and working condition correspondence, and pairs sound and vibration segments according to the same time interval and the same working condition to generate sound and vibration anomaly pairing feature data. The abnormal source competition discrimination layer, based on the sound and vibration abnormality pairing feature data, performs source pointing marking and competition discrimination according to the source type of the target device body abnormality, external sound source interference, external vibration impact, structural propagation abnormality and acquisition abnormality, and generates candidate source matching data; The working condition consistency screening layer, based on candidate source matching data, collaboratively determines the correspondence between abnormal sound segments, abnormal vibration segments, and working condition correspondence, and retains and merges the candidate sources determined to be consistent sound and vibration working conditions, and outputs consistent sound and vibration working condition abnormal source data.

[0011] As a preferred embodiment of the electrical equipment operating status evaluation method based on joint sound and vibration detection described in this invention, the specific steps for generating consistent abnormal source data of sound and vibration operating conditions are as follows: Based on the candidate source matching data, the corresponding sound abnormality segment markers, vibration abnormality segment markers, and working condition corresponding markers are extracted according to the source type of the candidate source to obtain the abnormal occurrence time and abnormal duration interval corresponding to each candidate source. When the sound abnormality segment marker, vibration abnormality segment marker, and operating condition corresponding marker all correspond to the same abnormality occurrence time and the same abnormality duration interval, the candidate source is determined to be a consistent source of sound and vibration operating conditions. When the sound abnormality segment marker, vibration abnormality segment marker, and operating condition corresponding marker do not simultaneously correspond to the same abnormality occurrence time and the same abnormality duration interval, the candidate source is determined to be an inconsistent source of sound and vibration operating conditions. Sources with consistent acoustic and vibration conditions are retained and merged, while sources with inconsistent acoustic and vibration conditions are removed from the candidate source matching data to generate data on anomalous sources with consistent acoustic and vibration conditions.

[0012] As a preferred embodiment of the electrical equipment operating status evaluation method based on joint sound and vibration detection described in this invention, the specific steps for generating sound and vibration energy conservation deviation data are as follows: Based on the consistent abnormal source data of acoustic and vibration working conditions, the corresponding relationship between the sound abnormal segment, vibration abnormal segment and working condition is extracted from each consistent abnormal source of acoustic and vibration working conditions. The energy change within the same abnormal duration interval of the sound abnormal segment and vibration abnormal segment is extracted to obtain the acoustic and vibration energy change data. Based on the acoustic and vibration energy change data, the sound energy change and vibration energy change are recombined according to the consistent abnormal source under the same acoustic and vibration working condition to obtain the corresponding acoustic and vibration energy data. Using the body's abnormal energy conversion benchmark, the deviation comparison of sound energy changes and vibration energy changes in the corresponding sound and vibration energy data is performed to determine whether the corresponding sound and vibration energy data belongs to the body energy consistent state, sound side deviation state, and vibration side deviation state. Based on the determination results, source classification is performed to form sound and vibration energy conservation deviation data.

[0013] As a preferred embodiment of the electrical equipment operating status evaluation method based on joint sound and vibration detection described in this invention, the specific steps for generating sound and vibration energy conservation deviation data are as follows: Based on the corresponding data of acoustic and vibration energy, extract the sound energy change, vibration energy change and working condition correspondence corresponding to the consistent abnormal source of each acoustic and vibration working condition, and match the body abnormal energy conversion benchmark under the corresponding working condition. When the correspondence between sound energy change and vibration energy change falls within the range corresponding to the abnormal energy conversion benchmark of the body, it is determined to be a state of consistent body energy. When the sound energy change deviates from the abnormal energy conversion benchmark of the body and the vibration energy change does not form a corresponding change, it is determined to be a state of sound side deviation. When the vibration energy change deviates from the abnormal energy conversion benchmark of the body and the sound energy change does not form a corresponding change, it is determined to be a state of vibration side deviation. By combining the vibration side deviation state with the source type label in the candidate source matching data, when the source type label corresponds to external vibration impact, it is classified as external vibration impact associated data; when the source type label corresponds to structural propagation anomaly, it is classified as structural propagation anomaly associated data. The classification results are then organized to obtain acoustic vibration energy conservation deviation data.

[0014] As a preferred embodiment of the electrical equipment operation status evaluation method based on joint sound and vibration detection according to the present invention, the specific steps for generating the target electrical equipment operation status evaluation result are as follows: Based on the acoustic and vibration energy conservation deviation data, the source type, energy deviation status and working condition correspondence in the acoustic and vibration working condition consistency anomaly source data are extracted. Ontology association verification and source marking are performed on each anomaly source to obtain source correction anomaly data. Based on the source correction of abnormal data, the data corresponding to external sound source interference, external vibration impact, structural propagation anomalies and acquisition anomalies are merged and organized to obtain non-body anomaly records. The anomaly sources that meet the body anomaly energy conversion benchmark are merged and organized to obtain body anomaly evaluation records. By utilizing non-physical anomaly records and physical anomaly evaluation records, the anomaly sources, acoustic and vibration deviation states, and corresponding operating conditions of the target electrical equipment are reorganized to generate the operating status evaluation results of the target electrical equipment.

[0015] As a preferred embodiment of the electrical equipment operating status evaluation method based on joint sound and vibration detection described in this invention, the specific steps for obtaining non-physical anomaly records and physical anomaly evaluation records are as follows: Based on the source correction of abnormal data, extract the source type, sound abnormal segment, vibration abnormal segment, working condition correspondence and sound and vibration energy conservation deviation data corresponding to each abnormal source, and classify the data corresponding to external sound source interference, external vibration impact, structural propagation abnormality and acquisition abnormality into non-body abnormality categories according to the source type. Data classified as non-entity anomalies are merged according to the time of occurrence of the anomaly, the duration of the anomaly, and the correspondence with the working conditions. Duplicate sound anomaly segments and vibration anomaly segments are deleted to obtain non-entity anomaly records. For anomaly sources that meet the energy conversion benchmark of the device body, the data on the target device body anomaly, anomaly occurrence time, anomaly duration interval, and acoustic and vibration energy conservation deviation are merged, and the corresponding sound anomaly segments, vibration anomaly segments, and working condition correspondence are associated and saved to obtain the body anomaly evaluation record.

[0016] The beneficial effects of this invention are as follows: by correcting the sources of anomalies through the deviation data of acoustic and vibration energy conservation, external sound source interference, external vibration impact, structural propagation anomalies and acquisition anomalies are classified into non-physical anomaly records, and physical anomaly sources are classified into physical anomaly evaluation records, so that the operation status evaluation results can more accurately reflect the physical anomaly changes of the target electrical equipment and improve the reliability of early warning. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method for evaluating the operating status of electrical equipment based on joint sound and vibration detection.

[0019] Figure 2 A flowchart for the formation of a set of acoustic and vibrational abnormal events.

[0020] Figure 3 This is a flowchart illustrating the competition and resolution relationship of acoustic vibration anomaly sources.

[0021] Figure 4 A flowchart for generating linkage based on the operational status evaluation results. Detailed Implementation

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0024] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0025] Reference Figures 1-4 As one embodiment of the present invention, this embodiment provides a method for evaluating the operating status of electrical equipment based on joint detection of sound and vibration, including the following steps: S1: Acquire the acoustic and vibration operating condition data and the abnormal energy conversion benchmark of the target electrical equipment, perform time alignment, operating condition segmentation and invalid acquisition segments on the acoustic and vibration operating condition data to form an acoustic and vibration synchronization segment.

[0026] S1.1: Extract the sound acquisition data and vibration acquisition data of the acoustic and vibration working conditions data, and perform time alignment and working condition segmentation to obtain segmented acoustic and vibration working condition data.

[0027] Furthermore, sound acquisition data, vibration acquisition data, and operating condition data are extracted from the sound and vibration condition data. The sound acquisition data and vibration acquisition data are sorted sequentially according to the acquisition time, and data points with duplicate acquisition times are deleted. The time of change of operating condition status in the operating condition data is used as the segment boundary, and the sound acquisition data and vibration acquisition data are mapped onto the same time axis. For data points in the sound acquisition data whose acquisition time is inconsistent with that in the vibration acquisition data, they are matched according to the nearest acquisition time, so that the sound segments and vibration segments within the same acquisition time period form a corresponding relationship. According to the start-up status, stable operation status, load change status, and shutdown status in the operating condition data, the sound segments and vibration segments that have completed time alignment are segmented and marked. The sound segments, vibration segments, and operating condition segments belonging to the same operating condition status and the same acquisition time period are associated and sorted to obtain segmented sound and vibration condition data.

[0028] It should be noted that when performing proximity matching on sound and vibration data, the maximum allowable time difference between the sound and vibration acquisition times is first determined. This maximum allowable time difference is determined by the sound sampling period, vibration sampling period, clock synchronization error of the acquisition equipment, and mechanical response delay of the target electrical equipment under the corresponding operating conditions. This maximum allowable time difference is used as the time synchronization tolerance. When the time difference between the sound and vibration acquisition times does not exceed the maximum allowable time difference, the data points are considered to be corresponding within the same acquisition period and proximity matching is performed. When the time difference exceeds the maximum allowable time difference, the correspondence between the sound segment and the vibration segment is no longer forcibly established. Instead, the corresponding data point or segment is marked as a time misalignment segment or asynchronous segment and is handled by invalid acquisition segment filtering to avoid data from different operating conditions or different abnormal stages being incorrectly bound.

[0029] S1.2: Using segmented acoustic and vibration operating condition data, invalid acquisition segments with sampling gaps, time misalignments, and acquisition interruptions are screened out, and the retained sound segments, vibration segments, operating condition segments, and body abnormal energy conversion benchmarks are correlated and recombined to form acoustic and vibration synchronization segments.

[0030] Furthermore, the corresponding sound and vibration segments are read according to the start and end times of the operating condition segments. The acquisition time of the sound segment, the acquisition time of the vibration segment, the time range of the operating condition segment, the sound sampling period, and the vibration sampling period are used as the verification objects. First, the missing rate judgment limit, the maximum allowable time misalignment, the continuous blank duration judgment limit, and the minimum number of effective sampling points are determined based on the sound sampling period, vibration sampling period, sensor refresh rate, communication buffer period, and operation and maintenance procedures. Within the same operating condition segment, when the actual number of sampling points of the sound segment or vibration segment is lower than the number of sampling points that should be sampled in the corresponding operating condition segment, and the missing rate reaches the missing rate judgment limit, the corresponding segment is marked as a sampling missing segment. When the difference between the acquisition time of the sound segment and the acquisition time of the vibration segment exceeds the maximum allowable time misalignment, or cannot fall within the time range of the same operating condition segment at the same time, the corresponding segment is marked as a time misalignment segment. When the duration of continuous absence of effective sampling data for the sound segment or vibration segment within the same operating condition segment reaches the continuous blank duration judgment limit, or the number of effective sampling points is lower than the minimum number of effective sampling points, the corresponding segment is marked as an acquisition interruption segment.

[0031] After filtering out missing sampling segments, time-displaced segments, and interrupted acquisition segments, data segments with consistent acquisition time, consistent operating conditions, and continuous existence of both sound and vibration segments are retained. The retained sound, vibration, and operating condition segments are grouped according to the same operating condition segment, and the correlation between the body abnormal energy conversion benchmark corresponding to the same operating condition segment and the grouped sound, vibration, and operating condition segments is established to form a sound-vibration synchronization segment.

[0032] Specifically, the abnormal energy conversion benchmark is formed from the historical acoustic and vibration synchronization segments in the historical operating data of the target electrical equipment. First, the sound abnormality segments, vibration abnormality segments and corresponding operating conditions that have been confirmed to belong to the abnormality of the target equipment are extracted from the historical acoustic and vibration synchronization segments. Then, the sound energy changes, vibration energy changes, the order of changes and the direction of changes under the same operating condition correspondence are merged and sorted to form the abnormal energy conversion benchmark corresponding to the operating condition correspondence.

[0033] The abnormal energy conversion benchmark of the target electrical equipment is formed by grouping the target electrical equipment according to its equipment model, installation location, load condition and speed range. Within each group, historical acoustic and vibration synchronous segments that have been confirmed to be abnormal in the target equipment are extracted, and the distribution range of corresponding sound energy changes, vibration energy changes, the order of changes and the direction of changes are statistically analyzed. The range of acoustic and vibration energy that appears continuously and stably within the group is used as the upper and lower limits of the abnormal energy conversion benchmark of the target equipment.

[0034] Each group shall have no fewer than thirty valid samples for forming the body's abnormal energy conversion benchmark. If the number of valid samples in a group is less than thirty, the calibration benchmark formed by the same model of equipment at the same installation location, under the same load conditions or similar speed range shall be used first. If this is still not enough, the factory test benchmark of the target electrical equipment shall be used as a temporary comparison benchmark. The acoustic and vibration energy deviation results obtained based on the temporary comparison benchmark shall be marked as pending verification and shall not be directly used as the final body abnormality evaluation record.

[0035] S2: Extract abnormal sound segments and abnormal vibration segments based on the acoustic-vibration synchronization segments, and associate and organize them according to the time of occurrence of abnormality, the duration of abnormality, and the corresponding relationship of working conditions to form a set of acoustic-vibration abnormal events.

[0036] S2.1: Extract the sound data segment and vibration data segment of the sound-vibration synchronization segment, and perform fluctuation interval identification and abnormal start and end location respectively to obtain the sound abnormal segment, vibration abnormal segment and the corresponding abnormal occurrence time.

[0037] Furthermore, the acoustic-vibration synchronization segment is read segment by segment according to the working condition segment. The sound data segment and vibration data segment arranged continuously according to the sampling time are extracted within the same working condition segment. DC offset elimination, short-time window division and envelope extraction are performed on the sound data segment and vibration data segment respectively to obtain the sound short-time window sequence and vibration short-time window sequence.

[0038] The short-time window length is determined based on the sound sampling period, vibration sampling period, and duration of a single mechanical action of the target electrical equipment. The window step size is smaller than the window length, and overlapping sampling intervals are retained between adjacent windows. The overlap rate is determined based on the sensor refresh rate and the rate of change of operating conditions.

[0039] Based on the sound short-time window sequence and the vibration short-time window sequence, the short-time energy, envelope peak value and the change amplitude of adjacent windows for each short-time window are calculated respectively. After removing windows marked as sampling missing segments, time misalignment segments and acquisition interruption segments from the same working condition segment, windows that exist continuously and whose short-time energy, envelope peak value and the change amplitude of adjacent windows all remain stable are selected as stable windows. The number of stable windows is not less than thirty, or not less than the number of windows contained in a complete stable operating cycle within the corresponding working condition segment.

[0040] For the short-time energy, envelope peak value, and variation amplitude of adjacent windows in the stable window, the mean, standard deviation, and quantile intervals are calculated respectively. The range formed by adding or subtracting two to three times the standard deviation from the mean, or the range from the low quantile to the high quantile of the stable sample, is used as the upper and lower boundaries of the normal fluctuation zone under the corresponding operating condition. When the short-time energy, envelope peak value, or variation amplitude of adjacent windows of a certain short-time window deviates from the normal fluctuation zone, it is first marked as a candidate abnormal window. Only when three or more consecutive candidate abnormal windows deviate from the normal fluctuation zone, or when the deviated state continuously covers a complete mechanical response interval, are the corresponding windows merged into an abnormal fluctuation interval to avoid single-point sampling noise triggering abnormal judgment.

[0041] Window segments that continuously deviate from the normal fluctuation band in the short-time window sequence of sound are merged into sound fluctuation intervals, and window segments that continuously deviate from the normal fluctuation band in the short-time window sequence of vibration are merged into vibration fluctuation intervals. The start and end points of anomalies are located for both sound and vibration fluctuation intervals. The starting sampling time of the short-time window that first deviates from the normal fluctuation band in each fluctuation interval is determined as the anomaly occurrence time, and the ending sampling time of the last short-time window before returning to the normal fluctuation band in each fluctuation interval is determined as the anomaly end time. Based on the anomaly occurrence time and the anomaly end time, corresponding data segments are extracted to obtain sound anomaly segments, vibration anomaly segments, and their corresponding anomaly occurrence times.

[0042] It should be noted that the normal fluctuation band is formed based on the continuous short-time sound window sequence and short-time vibration window sequence within the same working condition segment. First, data windows marked as missing sampling segments, time misaligned segments, and acquisition interruption segments are excluded. Then, the short-time energy, envelope peak value, and adjacent window variation amplitude in the remaining windows are extracted. The continuously and stably changing short-time energy, envelope peak value, and adjacent window variation amplitude within the same working condition segment are organized into the normal fluctuation band.

[0043] S2.2: Based on the abnormal sound segments, abnormal vibration segments and the corresponding occurrence time of the abnormality, match and organize the correspondence between the abnormal duration interval and the working condition to obtain the sound and vibration abnormality associated segments.

[0044] Furthermore, the sound and vibration anomaly segments are sequentially sorted according to their occurrence time. The duration of the sound anomaly is determined by the start and end times of each sound anomaly segment, and the duration of the vibration anomaly is determined by the start and end times of each vibration anomaly segment. The durations of the sound and vibration anomalies are then compared within the same acoustic-vibration synchronization segment. The corresponding operating condition segments for the sound and vibration anomalies are read. When the durations of the sound and vibration anomalies overlap in time and both correspond to the same operating condition segment, the sound anomaly segment, the vibration anomaly segment, the overlapping duration of the anomaly, the anomaly occurrence time, and the corresponding operating condition are bound together. The bound data is sorted according to the occurrence time of the anomaly, and the bound data that occur consecutively and are time-sequential within the same operating condition segment are merged. Duplicate sound and vibration anomaly segments are deleted to obtain the acoustic-vibration anomaly associated segments.

[0045] S2.3: Using the correlation segments of acoustic and vibration anomalies, the corresponding sound anomaly segments and vibration anomaly segments under the same working condition are recombined to form a set of acoustic and vibration anomaly events.

[0046] Furthermore, the sound and vibration anomaly segments are grouped according to the working condition correspondence to obtain sound anomaly segment groups and vibration anomaly segment groups under the same working condition. Then, the sound anomaly segment groups and vibration anomaly segment groups are sorted according to the time of anomaly occurrence. The anomaly duration interval of the sound anomaly segment is used as the matching benchmark to find vibration anomaly segments whose anomaly occurrence time falls within the same anomaly duration interval.

[0047] When one sound abnormality segment corresponds to multiple vibration abnormality segments, the vibration abnormality segment with the longest overlap time of the abnormal duration interval is selected as the corresponding vibration abnormality segment. When one vibration abnormality segment corresponds to multiple sound abnormality segments, the sound abnormality segment with the closest abnormality occurrence time is selected as the corresponding sound abnormality segment. The corresponding sound abnormality segment, vibration abnormality segment, abnormality occurrence time, abnormal duration interval and working condition correspondence are merged to form a single sound and vibration abnormality event.

[0048] For all corresponding sound and vibration abnormality segments under the same working condition, the merging process is repeated. For isolated sound or vibration abnormality segments that do not form a correspondence between sound and vibration abnormality segments, they are not merged as sound-vibration synchronous pairing events, but are retained as single-sided abnormality candidate segments. The occurrence time, duration interval, working condition correspondence, and segment source type are recorded. The merged sound and vibration abnormality events and the retained single-sided abnormality candidate segments are arranged in the order of occurrence time to form a set of sound and vibration abnormality events.

[0049] S3: Input the set of acoustic and vibration anomaly events into the acoustic and vibration anomaly source competition resolution model. According to the source type of the target equipment body anomaly, external sound source interference, external vibration impact, structural propagation anomaly, and acquisition anomaly, perform source matching and competition screening of sound anomaly segments, vibration anomaly segments, and corresponding working conditions to generate acoustic and vibration working condition consistent anomaly source data. The acoustic and vibration anomaly source competition resolution model is constructed based on the acoustic and vibration anomaly pairing identification layer, anomaly source competition discrimination layer, and working condition consistency screening layer.

[0050] S3.1: Acoustic and vibration anomaly pairing and identification layer. Based on the set of acoustic and vibration anomaly events, it organizes the features of the correspondence between sound anomaly segments, vibration anomaly segments, anomaly occurrence time, anomaly duration interval and working conditions, and pairs acoustic and vibration segments according to the same time interval and the same working condition to generate acoustic and vibration anomaly pairing feature data.

[0051] S3.2: Anomaly Source Competition Discrimination Layer. Based on the sound and vibration anomaly pairing feature data, it performs source pointing marking and competition discrimination according to the source type of the target equipment body anomaly, external sound source interference, external vibration impact, structural propagation anomaly and acquisition anomaly, and generates candidate source matching data.

[0052] S3.3: Working condition consistency screening layer. Based on candidate source matching data, it performs collaborative judgment on the correspondence of abnormal sound segments, abnormal vibration segments, and working condition correspondence, and retains and merges the candidate sources that are judged as consistent sound and vibration working conditions, and outputs the consistent sound and vibration working condition abnormal source data.

[0053] Based on the historical operating data of the target electrical equipment, historical acoustic and vibration synchronization segments that have been time-aligned, segmented by operating conditions, and filtered out by invalid acquisition segments are read. Historical abnormal sound segments, historical abnormal vibration segments, historical abnormal occurrence time, historical abnormal duration interval, and historical operating condition correspondence are extracted to form historical acoustic and vibration abnormal event samples. According to maintenance records, on-site verification records, operation logs, and acquisition abnormality records, the historical acoustic and vibration abnormal event samples are marked with source type to form training sample data with source type marking.

[0054] The training sample data with source type labels are hierarchically organized according to equipment model, operating condition category and source type, and divided into training samples, validation samples and test samples. From each historical acoustic and vibration anomaly event sample, the start and end time of the sound anomaly segment, the short-term energy of the sound, the peak value of the sound envelope, the start and end time of the vibration anomaly segment, the change of vibration energy, the duration of overlap of the acoustic and vibration anomaly duration interval, the operating condition label, the measurement point location label and the acquisition anomaly label are extracted to form the model input feature vector.

[0055] The model input feature vector is input into the acoustic vibration anomaly pairing identification layer, and the training output is historical acoustic vibration anomaly pairing feature data. The historical acoustic vibration anomaly pairing feature data and source type labels are input into the anomaly source competition discrimination layer, and the training output is each candidate source type and corresponding source confidence. The historical candidate source matching data is input into the working condition consistency filtering layer, and the training output is historical acoustic vibration working condition consistent sources and historical acoustic vibration working condition inconsistent sources.

[0056] During training, the source type identification error, anomaly occurrence time matching error, anomaly duration interval matching error, and operating condition consistency judgment error are used as joint training objectives to iteratively update the parameters of the acoustic and vibration anomaly pairing identification layer, anomaly source competition discrimination layer, and operating condition consistency screening layer. When the source type, anomaly occurrence time, anomaly duration interval, operating condition correspondence, and acoustic and vibration operating condition consistency judgment results in the validation and test samples all meet the qualification conditions, and the error change amplitude of multiple consecutive training rounds meets the convergence condition, the current training parameters are retained to obtain the trained acoustic and vibration anomaly source competition resolution model. For cases where there are multiple candidate sources for the same anomaly event, the candidate sources are competitively ranked according to source confidence, acoustic and vibration anomaly duration interval overlap, operating condition correspondence consistency, and collection anomaly label. The candidate source with the highest ranking and meeting the operating condition consistency requirement is retained as the model output result.

[0057] S3.4: Based on the set of acoustic and vibration abnormal events, the acoustic and vibration abnormal source competition resolution model is used to extract the corresponding sound abnormal segment, vibration abnormal segment, abnormal occurrence time, abnormal duration interval and working condition correspondence for each abnormal event. Sound abnormal segments and vibration abnormal segments that belong to the same abnormal occurrence time, the same abnormal duration interval and the same working condition correspondence are paired and organized to obtain acoustic and vibration abnormal pairing data.

[0058] Furthermore, based on the set of acoustic and vibration abnormal events, the acoustic and vibration abnormality source competition resolution model is used to read each abnormal event in chronological order of the abnormality occurrence time. From each abnormal event, the sound abnormal segment, vibration abnormal segment, abnormal occurrence time, abnormal duration interval, and working condition correspondence are extracted. Based on the working condition correspondence, abnormal events belonging to the same working condition correspondence are grouped into the same working condition event group. Within each working condition event group, the sound abnormal segment and vibration abnormal segment are arranged according to the abnormality occurrence time.

[0059] Based on the arranged sound and vibration anomaly segments, the occurrence time of each sound anomaly segment is compared with the occurrence time of each vibration anomaly segment. Then, the duration of the anomaly is compared for sound and vibration anomaly segments with the same occurrence time. When a sound and vibration anomaly segment simultaneously corresponds to the same occurrence time, the same duration, and the same operating condition, the sound anomaly segment, vibration anomaly segment, occurrence time, duration, and operating condition correspondence are merged into a set of sound-vibration anomaly pairing records. When a sound and vibration anomaly segment does not simultaneously correspond to the same occurrence time, duration, and operating condition, sound-vibration anomaly pairing record generation is not performed. All sound-vibration anomaly pairing records are then sorted sequentially according to the occurrence time and operating condition correspondence to obtain sound-vibration anomaly pairing data.

[0060] S3.5: Based on the sound and vibration anomaly pairing data, and focusing on the source types of the target equipment body anomaly, external sound source interference, external vibration impact, structural propagation anomaly, and acquisition anomaly, perform source pointing marking and correspondence verification for each set of sound and vibration anomaly pairing data to obtain candidate source matching data.

[0061] Furthermore, based on the sound and vibration anomaly pairing data, the overlap between the sound anomaly segment and the vibration anomaly segment in the time of anomaly occurrence and the duration of anomaly is first checked. When there is overlap in the same time interval and they belong to the same working condition, the corresponding sound and vibration anomaly pairing data is entered into the source pointing mark. Then, the sound and vibration anomaly pairing data that has entered the source pointing mark processing is classified into source types.

[0062] For the single-sided anomalous candidate segments retained in the set of acoustic and vibration anomalous events, they are processed into source pointing markers according to the segment source type, anomalous occurrence time, anomalous duration interval, and working condition correspondence. When only acoustic anomalous segments exist and vibration anomalous segments lack corresponding changes, they are considered as candidate evidence of external sound source interference or acquisition anomalies. When only vibration anomalous segments exist and acoustic anomalous segments lack corresponding changes, they are considered as candidate evidence of external vibration impact, structural propagation anomalies, or acquisition anomalies.

[0063] When abnormal sound segments and abnormal vibration segments occur simultaneously under the same operating condition and the measuring point points to the target electrical equipment, it is marked as an abnormality of the target equipment itself. When an abnormal sound segment exists but the abnormal vibration segment lacks a corresponding change, it is marked as external sound source interference. When an abnormal vibration segment exists but the abnormal sound segment lacks a corresponding change, it is marked as external vibration impact. When both abnormal sound segments and abnormal vibration segments exist but there is a propagation misalignment between the measuring point position and the abnormal duration interval, it is marked as structural propagation abnormality. When the corresponding sampling of an abnormal sound segment or abnormal vibration segment is missing, time is misaligned, or acquisition is interrupted, it is marked as acquisition abnormality.

[0064] The corresponding relationships of the sound and vibration anomaly pairing data with completed source pointing marks are verified. The correspondence relationships of sound anomaly segments, vibration anomaly segments, and operating conditions are checked to see if they are all true at the same time. The source type, sound anomaly segment correspondence, vibration anomaly segment correspondence, operating condition correspondence, anomaly occurrence time, and anomaly duration interval are sorted out to obtain candidate source matching data.

[0065] It should be noted that whether the conditions are simultaneously met is determined when the abnormal sound segments and abnormal vibration segments under the same candidate source correspond to the same abnormal occurrence time or have overlapping abnormal duration intervals, and both fall into the same working condition correspondence relationship. At the same time, the abnormal sound segments and abnormal vibration segments can be explained by the source type of the candidate source. If the abnormal sound segments, abnormal vibration segments, and working condition correspondence relationships can correspond to each other in terms of time interval, working condition attribution, and source type, then the abnormal sound segment correspondence relationship, abnormal vibration segment correspondence relationship, and working condition correspondence relationship are determined to be simultaneously met.

[0066] S3.6: Based on the candidate source matching data, extract the corresponding sound abnormality segment markers, vibration abnormality segment markers, and operating condition corresponding markers according to the source type of the candidate source, and obtain the abnormal occurrence time and abnormal duration interval corresponding to each candidate source.

[0067] Furthermore, according to the source types of target equipment anomalies, external sound source interference, external vibration impact, structural propagation anomalies, and acquisition anomalies, the candidate source matching data are classified and arranged, and candidate records belonging to the same candidate source are located one by one under each source type. From the candidate records of the same candidate source, sound anomaly segment markers, vibration anomaly segment markers, and operating condition corresponding markers are read.

[0068] Sound anomaly segment markers are used to identify the start, end, and segment number of a corresponding sound anomaly segment. Vibration anomaly segment markers are used to identify the start, end, and segment number of a corresponding vibration anomaly segment. Working condition correspondence markers are used to identify the start, end, and working condition category of a working condition segment in the set of sound and vibration anomaly events. The start time of the sound anomaly segment corresponding to the sound anomaly segment marker, the start time of the vibration anomaly segment corresponding to the vibration anomaly segment marker, and the start time of the working condition segment corresponding to the working condition correspondence marker are merged. The earliest anomaly start time that falls into the same working condition segment among the three is determined as the anomaly occurrence time corresponding to each candidate source.

[0069] The boundaries of the sound abnormality segment endpoint, the vibration abnormality segment endpoint, and the working condition segment endpoint are sorted out according to the sound abnormality segment marker, the vibration abnormality segment endpoint, and the working condition segment endpoint. The time range that the sound abnormality segment and the vibration abnormality segment jointly cover within the same working condition segment is determined as the abnormality duration interval corresponding to each candidate source. The source type, sound abnormality segment marker, vibration abnormality segment marker, working condition corresponding marker, abnormality occurrence time, and abnormality duration interval are associated and sorted according to the same candidate source to obtain the abnormality occurrence time and abnormality duration interval corresponding to each candidate source.

[0070] S3.7: When the sound abnormality segment marker, vibration abnormality segment marker, and operating condition corresponding marker all correspond to the same abnormality occurrence time and the same abnormality duration interval, the candidate source is determined to be a consistent source of sound and vibration operating conditions. When the sound abnormality segment marker, vibration abnormality segment marker, and operating condition corresponding marker do not simultaneously correspond to the same abnormality occurrence time and the same abnormality duration interval, the candidate source is determined to be an inconsistent source of sound and vibration operating conditions.

[0071] Furthermore, the candidate source matching data are grouped and organized according to the source type of the candidate source, and the abnormal sound segment marker, abnormal vibration segment marker, working condition corresponding marker, abnormal occurrence time and abnormal duration interval are extracted from each group of candidate source matching data. The abnormal occurrence time and abnormal duration interval are used as the same source determination criteria.

[0072] The abnormal occurrence time corresponding to the sound abnormal segment marker, the abnormal occurrence time corresponding to the vibration abnormal segment marker, and the working condition segment time corresponding to the working condition corresponding marker are compared item by item. Then, the abnormal duration interval corresponding to the sound abnormal segment marker, the abnormal duration interval corresponding to the vibration abnormal segment marker, and the working condition duration interval corresponding to the working condition corresponding marker are compared item by item. When the sound abnormal segment marker, vibration abnormal segment marker, and working condition corresponding marker all exist under the same candidate source, and the difference between the abnormal start time of the sound abnormal segment and the vibration abnormal segment does not exceed the time tolerance threshold, and the overlap ratio of the abnormal duration interval is not lower than the interval overlap judgment threshold, and the sound abnormal segment and the vibration abnormal segment both fall within the same working condition segment range, the candidate source is judged as a consistent source of sound and vibration working conditions. When the sound abnormal segment marker, vibration abnormal segment marker, or working condition corresponding marker is missing, or the difference between the abnormal start time exceeds the time tolerance threshold, or the overlap ratio of the abnormal duration interval is lower than the interval overlap judgment threshold, or the sound abnormal segment and the vibration abnormal segment do not fall within the same working condition segment range, the candidate source is judged as an inconsistent source of sound and vibration working conditions.

[0073] When there are missing sound abnormality segment markers, vibration abnormality segment markers, or corresponding operating condition markers under the same candidate source, or when the abnormality occurrence times corresponding to the three types of markers are inconsistent, or when the abnormality duration intervals corresponding to the three types of markers are inconsistent, the candidate source will be determined as a source of inconsistent sound and vibration operating conditions.

[0074] The formula for determining the consistent source of acoustic and vibration conditions is: ; ; in, Indicates the first Anomalous acoustic events in candidate sources The consistent judgment value of acoustic and vibration conditions. This represents the event number in the set of acoustic and vibrational abnormal events. This indicates the type of candidate source, including anomalies in the target device itself, interference from external sound sources, external vibration and shock, structural propagation anomalies, and acquisition anomalies. Indicates candidate source Whether the consistent acoustic and vibration conditions are met is indicated by a value of 1, which indicates that the conditions are met, and a value of 0 indicates that the conditions are not met. This indicates the duration of overlap between abnormal sound segments and abnormal vibration segments within the abnormal duration interval, measured in units of time. >0 indicates that the two events overlap in time. This indicates that the abnormal sound segment belongs to the operating condition. This indicates that the abnormal vibration segment belongs to a working condition; both are working condition category markers and are dimensionless. Indicates candidate source Can you explain the abnormal audio segments? Indicates candidate source Whether the vibration anomaly segment can be explained is a binary label; a value of 1 indicates that it can be explained, and a value of 0 indicates that it cannot be explained. This represents a conditional function. The value is 1 if all conditions within the parentheses are true, and 0 otherwise.

[0075] In the formula for determining the consistent source of acoustic and vibration operating conditions, the time quantity is only compared with the time quantity, the operating condition category is only judged for category consistency, and the source interpretation mark is only judged for binary value.

[0076] It should be noted that the time tolerance threshold is determined based on the sound sampling period, vibration sampling period, clock synchronization error of the acquisition device, and mechanical response delay of the target electrical equipment under the corresponding operating conditions. The value range is between one sampling period and one complete mechanical response delay period. The interval overlap judgment threshold is determined based on the statistical ratio of the overlap between the sound abnormality duration interval and the vibration abnormality duration interval of the confirmed abnormal sample in the historical sound and vibration synchronization segments of the same type of target electrical equipment. The value range is from 0 to 1. The closer the value is to 1, the higher the degree of overlap between the duration intervals of the sound abnormal segment and the vibration abnormal segment.

[0077] S3.8: Retain and merge sources with consistent acoustic and vibration conditions, and remove sources with inconsistent acoustic and vibration conditions from the candidate source matching data to generate data on anomalous sources with consistent acoustic and vibration conditions.

[0078] Furthermore, the source type, sound abnormal segment marker, vibration abnormal segment marker, working condition correspondence marker, abnormal occurrence time, and abnormal duration interval corresponding to each candidate source are read. Based on the collaborative judgment results of the sound abnormal segment correspondence, vibration abnormal segment correspondence, and working condition correspondence, the candidate sources are divided into sources with consistent sound and vibration working conditions and sources with inconsistent sound and vibration working conditions.

[0079] For candidate sources identified as consistent sources of acoustic and vibration operating conditions, the source type, abnormal sound segments, abnormal vibration segments, operating condition correspondence, abnormal occurrence time, and abnormal duration interval are retained. These are then merged according to the same source type, same abnormal occurrence time, same abnormal duration interval, and same operating condition correspondence. Duplicate abnormal sound segment and abnormal vibration segment markers are deleted. Abnormal sound segments and abnormal vibration segments belonging to the same abnormal duration interval are grouped into a single consistent source record for acoustic and vibration operating conditions. For candidate sources identified as inconsistent sources of acoustic and vibration operating conditions, the corresponding source type marker, abnormal sound segment marker, abnormal vibration segment marker, and operating condition correspondence marker in the candidate source matching data are deleted. These are no longer involved in the merging process of consistent source records for acoustic and vibration operating conditions. The merged consistent source records are arranged in chronological order of abnormal occurrence time. Each consistent source record is then associated with its corresponding abnormal sound segment, abnormal vibration segment, and operating condition correspondence to generate consistent abnormal source data for acoustic and vibration operating conditions.

[0080] S4: Based on the consistent anomaly source data of each acoustic and vibration working condition, extract the sound energy change and vibration energy change corresponding to each consistent anomaly source of each acoustic and vibration working condition, and compare the correspondence between the sound energy change and vibration energy change with the body's abnormal energy conversion benchmark to form acoustic and vibration energy conservation deviation data.

[0081] S4.1: Based on the consistent abnormal source data of acoustic and vibration working conditions, extract the sound abnormal segment, vibration abnormal segment and working condition correspondence in each consistent abnormal source of acoustic and vibration working conditions, extract the energy change within the same abnormal duration interval of the sound abnormal segment and vibration abnormal segment, and obtain acoustic and vibration energy change data.

[0082] Furthermore, based on the occurrence time and duration interval of the abnormality corresponding to the consistent abnormality source for each acoustic and vibration condition, and using the same duration interval as a common truncation boundary, the acoustic abnormality segment and the vibration abnormality segment are truncated separately. The truncated acoustic abnormality segment and vibration abnormality segment are then subjected to DC offset removal, amplitude continuity check and sampling length alignment to obtain acoustic and vibration segment data within the same interval.

[0083] Using acoustic and vibration segment data within the same interval, the amplitude fluctuations, energy accumulation changes, and energy differences before and after the interval are extracted for both sound and vibration anomalies within the same anomaly duration interval. The extracted data are then correlated and organized according to the working condition correspondence to obtain the sound energy changes and vibration energy changes under the same acoustic and vibration working condition with consistent anomaly sources. Based on the sound energy changes, vibration energy changes, anomaly occurrence time, anomaly duration interval, and working condition correspondence under each acoustic and vibration working condition with consistent anomaly sources, the data on acoustic and vibration energy changes are merged and organized to obtain the acoustic and vibration energy change data.

[0084] For example, if the target electrical equipment is a contactor in a distribution cabinet, and the source of the abnormal sound and vibration condition during a single engagement operation has been determined to be an abnormality in the target equipment itself, the time of the abnormality is 10:15:23, the duration of the abnormality is from 10:15:23 to 10:15:25, and the corresponding operating condition is the contactor engagement condition.

[0085] Within the abnormal duration range, the impact sound amplitude fluctuations at the moment of engagement, the cumulative changes in continuous noise energy, and the sound energy differences before and after the abnormal range are extracted from the sound abnormal segments. The vibration peak fluctuations, the cumulative changes in vibration energy, and the vibration energy differences before and after the abnormal range are extracted from the vibration abnormal segments. When merging and organizing, the sound energy changes and vibration energy changes under the same abnormal occurrence time, the same abnormal duration range, the same working condition, and the same abnormal source are placed into the same sound and vibration energy change record. The sound segments and vibration segments that are repeatedly collected or marked are deleted, and only the sound energy change data and vibration energy change data corresponding to the same contactor engagement abnormal process are retained to form sound and vibration energy change data under the same abnormal source with consistent sound and vibration working conditions.

[0086] S4.2: Based on the acoustic and vibration energy change data, the sound energy change and vibration energy change are recombined according to the consistent abnormal source under the same acoustic and vibration working condition to obtain the corresponding acoustic and vibration energy data.

[0087] Furthermore, for each audio-visual energy change data point, the corresponding audio-visual working condition consistent anomaly source, sound anomaly segment, vibration anomaly segment, anomaly occurrence time, anomaly duration interval, and working condition correspondence are extracted. Using the audio-visual working condition consistent anomaly source as the aggregation object, the sound energy changes and vibration energy changes belonging to the same audio-visual working condition consistent anomaly source and within the same anomaly duration interval are grouped into the same processing group.

[0088] Next, the changes in sound energy and vibration energy within the same treatment group are sorted according to the time of anomaly occurrence. The parts with the same working condition correspondence and overlapping anomaly duration intervals are taken as the corresponding intervals of sound and vibration energy. The sound energy changes and vibration energy changes within the corresponding intervals of sound and vibration energy are established. The sound energy changes and vibration energy changes with duplicate correspondences within the same treatment group are merged. The sound energy changes and vibration energy changes with complete anomaly duration interval coverage, measurement point locations consistent with sound and vibration working conditions, and consistent anomaly sources are retained. Energy change data with non-overlapping anomaly duration intervals or inconsistent working condition correspondences are deleted. The sound energy changes, vibration energy changes, sound and vibration energy corresponding intervals, and working condition correspondences corresponding to the consistent anomaly sources of each sound and vibration working condition are associated and organized to obtain the sound and vibration energy corresponding data.

[0089] S4.3: Based on the corresponding acoustic and vibration energy data, extract the sound energy change, vibration energy change and working condition correspondence corresponding to the consistent abnormal source of each acoustic and vibration working condition, and match the body abnormal energy conversion benchmark under the corresponding working condition.

[0090] Furthermore, according to the source type, occurrence time and duration of the consistent acoustic and vibration working conditions, the corresponding data of acoustic and vibration energy are sorted out item by item. The sound energy changes and vibration energy changes belonging to the same consistent acoustic and vibration working conditions are grouped into the same energy correspondence group, and the working condition correspondence relationship corresponding to the same energy correspondence group is extracted.

[0091] Using the working condition correspondence as the matching basis, the reference content corresponding to the current working condition of the target electrical equipment is found from the already acquired abnormal energy conversion reference of the body, and the found reference content is associated with the sound energy change and vibration energy change in the same energy correspondence group respectively.

[0092] For each consistent source of acoustic and vibration abnormality, the energy correspondence group is repeatedly sorted, the working condition correspondence is extracted, and the body abnormal energy conversion benchmark is matched to form energy comparison data including sound energy change, vibration energy change, working condition correspondence, and the matched body abnormal energy conversion benchmark.

[0093] S4.4: When the correspondence between the change in sound energy and the change in vibration energy falls within the range corresponding to the abnormal energy conversion benchmark of the body, it is determined to be a state of consistent body energy. When the change in sound energy deviates from the abnormal energy conversion benchmark of the body and the change in vibration energy does not form a corresponding change, it is determined to be a state of sound side deviation. When the change in vibration energy deviates from the abnormal energy conversion benchmark of the body and the change in sound energy does not form a corresponding change, it is determined to be a state of vibration side deviation.

[0094] Furthermore, the corresponding working condition relationship, abnormal duration interval, sound energy change, and vibration energy change for each consistent source of acoustic and vibration working conditions are extracted. The benchmark content that matches the working condition relationship is read from the body abnormal energy conversion benchmark. Based on the abnormal duration interval, the start point, peak point, and end position of sound energy change and vibration energy change are aligned to form a correspondence between sound energy change and vibration energy change. When the correspondence between sound energy change and vibration energy change is consistent with the body abnormal energy conversion benchmark, it is determined to be a consistent body energy state. When the sound energy change deviates from the body abnormal energy conversion benchmark and the vibration energy change does not form a corresponding change, it is determined to be a sound side deviation state. When the vibration energy change deviates from the body abnormal energy conversion benchmark and the sound energy change does not form a corresponding change, it is determined to be a vibration side deviation state.

[0095] The obtained states of consistent body energy, sound deviation, and vibration deviation are recorded as the state determination content of the acoustic and vibration energy conservation deviation data.

[0096] The formula for determining the deviation state of acoustic vibration energy is: ; in, Indicates the first Anomalous acoustic events in candidate sources The acoustic vibration energy deviation state under the condition, and the candidate sources For consistent acoustic and vibration operating conditions, Indicates a state of consistent energy across the entire entity. Indicates the state of sound deviation. Indicates the deviation state of the vibration side. Indicates the first The corresponding operating conditions for each acoustic and vibration anomaly event. This represents the ratio of energy changes obtained from changes in sound energy and vibration energy; it is a dimensionless quantity. and These represent the corresponding working conditions. The lower and upper limits of the energy change ratio in the lower body's abnormal energy transformation benchmark are both dimensionless quantities. This represents the time difference between the peak position of sound energy change and the peak position of vibration energy change, with the dimension of time. and These represent the corresponding working conditions. The lower and upper limits of the peak time difference in the lower body's abnormal energy conversion benchmark are both measured in units of time. This represents the normalized deviation of the sound energy change from the reference value of the sound energy change in the body's abnormal energy conversion reference. This represents the normalized deviation of the energy change on the vibration side from the benchmark value of the vibration energy change in the body's abnormal energy conversion benchmark. Both are dimensionless quantities and can be directly compared in magnitude.

[0097] S4.5: Combine the vibration side deviation state with the source type label in the candidate source matching data. When the source type label corresponds to external vibration impact, it is classified as external vibration impact associated data. When the source type label corresponds to structural propagation anomaly, it is classified as structural propagation anomaly associated data. The classification results are then sorted to obtain acoustic vibration energy conservation deviation data.

[0098] Furthermore, the acoustic and vibration energy corresponding data of the state determined to be a vibration side deviation are extracted from the deviation judgment results. According to the correspondence between the abnormal occurrence time, the abnormal duration interval, and the working condition, the candidate source record corresponding to the vibration side deviation state is searched in the candidate source matching data. The source type mark in the candidate source record is read. When the source type is marked as external vibration impact, the corresponding vibration side deviation state, sound energy change, vibration energy change, abnormal occurrence time, abnormal duration interval, and working condition correspondence are included in the external vibration impact associated data. When the source type is marked as structural propagation anomaly, the corresponding vibration side deviation state, sound energy change, vibration energy change, abnormal occurrence time, abnormal duration interval, and working condition correspondence are included in the structural propagation anomaly associated data.

[0099] The external vibration and impact correlation data and the structural propagation anomaly correlation data were merged and sorted according to the anomaly occurrence time and anomaly duration interval, respectively. Duplicate corresponding sound energy change and vibration energy change records were deleted. The sorted external vibration and impact correlation data, structural propagation anomaly correlation data, data corresponding to the body energy consistency state and data corresponding to the sound side deviation state were uniformly summarized to obtain sound and vibration energy conservation deviation data.

[0100] S5: Based on the acoustic and vibration energy conservation deviation data, the source correction is performed on the acoustic and vibration condition consistent anomaly source data. External sound source interference, external vibration impact, structural propagation anomaly and acquisition anomaly are classified into non-body anomaly records. Anomaly sources that meet the body anomaly energy conversion benchmark are classified into body anomaly evaluation records. The target electrical equipment operation status evaluation results are generated based on the non-body anomaly records and body anomaly evaluation records.

[0101] S5.1: Based on the acoustic and vibration energy conservation deviation data, extract the source type, energy deviation status and working condition correspondence from the acoustic and vibration working condition consistent anomaly source data, perform ontology association verification and source marking for each anomaly source, and obtain source correction anomaly data.

[0102] Furthermore, the source type, energy deviation state, working condition correspondence, anomaly occurrence time, and anomaly duration range of each consistent acoustic and vibration working condition anomaly source are extracted. The extracted content is then matched with the sound anomaly segments and vibration anomaly segments in the consistent acoustic and vibration working condition anomaly source data to form anomaly source data to be verified. Based on the anomaly source data to be verified, a body association verification is performed on each anomaly source. When the energy deviation state is a body energy consistency state, the source type is a target device body anomaly, and the working condition correspondence is consistent with the working condition to which the sound anomaly segment and vibration anomaly segment belong, the corresponding anomaly source is marked as a body association source. When the energy deviation state is a sound side deviation state, the anomaly source with the source type of external sound source interference is marked as an external sound source interference source, the anomaly source with the source type of acquisition anomaly is marked as an acquisition anomaly source, and the anomaly source with the source type of target device body anomaly but only sound side deviation is corrected to an external sound source interference source.

[0103] When the energy deviation state is a vibration side deviation state, the abnormal source with the source type of external vibration impact is marked as an external vibration impact source, the abnormal source with the source type of structural propagation abnormality is marked as a structural propagation abnormality source, and the abnormal source with the source type of target equipment body abnormality but only vibration side deviation is corrected to external vibration impact source or structural propagation abnormality source. After the source marking is completed, the body-related sources, external sound source interference sources, external vibration impact sources, structural propagation abnormality sources and acquisition abnormality sources are sorted and organized according to the abnormal occurrence time, abnormal duration interval and working condition correspondence. The source type, energy deviation state, source marking, sound abnormality segment and vibration abnormality segment are associated and saved to obtain source correction abnormality data.

[0104] Specifically, for anomalies originating from the target equipment itself but only exhibiting lateral vibration deviation, if the arrival time of vibrations at multiple measurement points shows transmission from the external contact location towards the target equipment, with the vibration amplitude decreasing along the transmission direction, and accompanied by synchronous vibrations or impact pulses from nearby equipment, then it is corrected to an external vibration impact source. If the arrival time and amplitude attenuation relationship of vibrations at multiple measurement points conform to the structural propagation path of the target equipment's support structure, cabinet connectors, or installation foundation, and no independent external impact pulses are formed, then it is corrected to a structural propagation anomaly source.

[0105] S5.2: Correct abnormal data based on the source, extract the source type, sound abnormal segment, vibration abnormal segment, working condition correspondence and sound and vibration energy conservation deviation data corresponding to each abnormal source, and classify the data corresponding to external sound source interference, external vibration impact, structural propagation abnormality and acquisition abnormality into non-body abnormality categories according to the source type.

[0106] Furthermore, data items corresponding to each source of anomaly are read one by one according to the time of occurrence of the anomaly and the duration of the anomaly. The source type, sound anomaly segment, vibration anomaly segment, working condition correspondence, and sound and vibration energy conservation deviation data are extracted from the data items corresponding to each source of anomaly.

[0107] The source types are matched one by one with external sound source interference, external vibration impact, structural propagation anomaly, and acquisition anomaly. When the source type is external sound source interference, the corresponding sound anomaly segments, vibration anomaly segments, working condition correspondence, and acoustic-vibration energy conservation deviation data are classified into the external sound source interference group in the non-body anomaly category. When the source type is external vibration impact, the corresponding sound anomaly segments, vibration anomaly segments, working condition correspondence, and acoustic-vibration energy conservation deviation data are classified into the external vibration impact group in the non-body anomaly category. When the source type is structural propagation anomaly, the corresponding sound anomaly segments, vibration anomaly segments, working condition correspondence, and acoustic-vibration energy conservation deviation data are classified into the structural propagation anomaly group in the non-body anomaly category. When the source type is acquisition anomaly, the corresponding sound anomaly segments, vibration anomaly segments, working condition correspondence, and acoustic-vibration energy conservation deviation data are classified into the acquisition anomaly group in the non-body anomaly category.

[0108] Data grouped into the same category are arranged according to the time of occurrence of anomalies and the duration of anomalies. Repeated sound and vibration anomaly segments with the same time of occurrence, the same duration of anomalies, and the same working conditions are merged to classify the data corresponding to external sound source interference, external vibration impact, structural propagation anomalies, and acquisition anomalies, thereby obtaining non-entity anomaly categories.

[0109] S5.3: Merge the data classified as non-subject anomalies according to the time of occurrence of the anomaly, the duration of the anomaly, and the corresponding working conditions, and delete duplicate corresponding sound anomaly segments and vibration anomaly segments to obtain non-subject anomaly records.

[0110] Furthermore, the source type, anomaly occurrence time, anomaly duration interval, working condition correspondence, sound anomaly segment, and vibration anomaly segment are extracted from the data classified as non-ontology anomalies. The data are then sorted according to source type, working condition correspondence, and anomaly occurrence time to obtain sorted non-ontology anomaly data. Based on the sorted non-ontology anomaly data, data with the same source type, the same working condition correspondence, and overlapping or consecutive anomaly duration intervals are grouped together to obtain non-ontology anomaly grouped data.

[0111] Based on the non-entity anomaly grouping data, the earliest anomaly occurrence time within the same group is taken as the merged anomaly occurrence time, and the start and end times covered within the same group are taken as the merged anomaly duration interval. Sound and vibration anomaly segments within the same group are then spliced ​​and organized according to the acquisition time order to obtain merged non-entity anomaly data. Using this merged data, duplicate segments are identified for data with the same anomaly occurrence time, anomaly duration interval, and operating condition correspondence, and where sound or vibration anomaly segments are repeated. The first occurrence of sound and vibration anomaly segments is retained, while duplicate corresponding sound and vibration anomaly segments are deleted, resulting in deduplicated non-entity anomaly data. This deduplicated non-entity anomaly data is then organized according to the source type, anomaly occurrence time, anomaly duration interval, operating condition correspondence, and the order of sound and vibration anomaly segments to obtain non-entity anomaly records.

[0112] S5.4: For the sources of anomalies that meet the energy conversion benchmark of the main body, merge them according to the target equipment body anomaly, the time of anomaly occurrence, the duration of anomaly, and the deviation data of acoustic and vibration energy conservation, and save the corresponding sound anomaly segments, vibration anomaly segments, and working condition correspondence to obtain the body anomaly evaluation record.

[0113] Furthermore, anomaly sources that meet the energy conversion benchmark of the body are screened from the source correction anomaly data, and the target equipment body anomaly, anomaly occurrence time, anomaly duration interval, acoustic and vibration energy conservation deviation data, sound anomaly segments, vibration anomaly segments and working condition correspondences corresponding to the anomaly sources that meet the target equipment body anomaly are extracted. Based on the target equipment body anomaly, the anomaly sources that meet the energy conversion benchmark of the body are categorized and grouped. Under the same target equipment body anomaly category, the anomaly sources are arranged in the order of anomaly occurrence time. Anomaly sources with overlapping anomaly duration intervals or consecutive beginning and end are merged into the same body anomaly evaluation unit.

[0114] Based on the acoustic and vibration energy conservation deviation data within the same body anomaly evaluation unit, the corresponding states of sound energy change and vibration energy change are summarized and organized. The acoustic and vibration deviation states, anomaly start time, anomaly end time and working condition correspondence within the same body anomaly evaluation unit are retained. Using the anomaly occurrence time and anomaly duration interval corresponding to the same body anomaly evaluation unit, the time position of sound anomaly segments and vibration anomaly segments is calibrated. The sound anomaly segments, vibration anomaly segments and working condition correspondences belonging to the same body anomaly evaluation unit are associated and recorded. The body anomaly evaluation units that have completed the association record are organized according to the target equipment body anomaly category. Duplicate records of sound anomaly segments, vibration anomaly segments and working condition correspondences are deleted to obtain the body anomaly evaluation record.

[0115] S5.5: Using non-physical anomaly records and physical anomaly evaluation records, the anomaly sources, acoustic and vibration deviation states, and corresponding operating conditions of the target electrical equipment are reorganized to generate the operating status evaluation results of the target electrical equipment.

[0116] Furthermore, the system is organized sequentially according to the working condition correspondence and the time of anomaly occurrence to obtain the system anomaly evaluation sequence. From the non-system anomaly records, the system extracts the corresponding sound anomaly segments, vibration anomaly segments, anomaly occurrence time, anomaly duration interval, sound and vibration deviation status, and working condition correspondence for external sound source interference, external vibration impact, structural propagation anomalies, and acquisition anomalies. These are then merged and organized according to source type, working condition correspondence, and anomaly occurrence time to obtain the non-system anomaly exclusion sequence. The system anomaly evaluation sequence and the non-system anomaly exclusion sequence are then checked against each other according to the anomaly occurrence time and anomaly duration interval. Sound and vibration anomaly segments that have already been included in the non-system anomaly exclusion sequence within the same anomaly duration interval are no longer included in the target equipment system anomaly evaluation scope. For sound and vibration anomaly segments that have been included in the system anomaly evaluation sequence within the same anomaly duration interval, the corresponding sound and vibration deviation status and working condition correspondence are retained.

[0117] The retained body anomaly evaluation sequence is associated and recombined with the non-body anomaly exclusion sequence so that each anomaly duration interval corresponds to a clear anomaly source, acoustic and vibration deviation state and working condition correspondence, and the target electrical equipment operating status evaluation result is formed according to the anomaly occurrence time.

[0118] Specifically, the target electrical equipment operation status evaluation result refers to the equipment status judgment result formed after sorting out the current operation status, abnormality source type, sound and vibration deviation status and corresponding working conditions of the target electrical equipment based on the physical abnormality evaluation record and non-physical abnormality record.

[0119] In summary, this invention corrects the sources of anomalies by using acoustic and vibration energy conservation deviation data, classifying external sound source interference, external vibration impact, structural propagation anomalies, and acquisition anomalies into non-physical anomaly records, and classifying physical anomaly sources into physical anomaly evaluation records. This makes the operational status evaluation results more accurately reflect the physical anomaly changes of the target electrical equipment and improves the reliability of early warning.

[0120] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for evaluating the operating status of electrical equipment based on joint detection of sound and vibration, characterized in that, include: Acquire acoustic and vibration condition data and abnormal energy conversion benchmark of the target electrical equipment, perform time alignment, condition segmentation and invalid acquisition segment screening on the acoustic and vibration condition data to form acoustic and vibration synchronization segments; Based on the acoustic-vibration synchronization segment, abnormal sound segments and abnormal vibration segments are extracted, and then correlated and organized according to the time of occurrence of abnormality, the duration of abnormality, and the corresponding working conditions to form a set of acoustic-vibration abnormal events; The set of acoustic and vibration abnormal events is input into the acoustic and vibration abnormality source competition resolution model. According to the source type of the target equipment body abnormality, external sound source interference, external vibration impact, structural propagation abnormality and acquisition abnormality, the source matching and competition screening of sound abnormality segments, vibration abnormality segments and working condition correspondence are performed to generate acoustic and vibration working condition consistent abnormality source data. Based on the consistent anomaly source data of acoustic and vibration working conditions, the sound energy change and vibration energy change corresponding to each consistent anomaly source of acoustic and vibration working conditions are extracted, and the correspondence between the sound energy change and vibration energy change is compared with the body anomaly energy conversion benchmark to form acoustic and vibration energy conservation deviation data. Based on the acoustic and vibration energy conservation deviation data, the source correction of the consistent abnormal source data of acoustic and vibration operating conditions is performed. External sound source interference, external vibration impact, structural propagation abnormality and acquisition abnormality are classified into non-body abnormality records. Abnormal sources that meet the body abnormality energy conversion benchmark are classified into body abnormality evaluation records. The target electrical equipment operation status evaluation results are generated based on the non-body abnormality records and body abnormality evaluation records.

2. The method for evaluating the operating status of electrical equipment based on joint detection of sound and vibration as described in claim 1, characterized in that, The specific steps for forming the acoustic-vibration synchronization segment are as follows: The sound and vibration data of the acoustic and vibration conditions are extracted, and time alignment and condition segmentation are performed to obtain segmented acoustic and vibration conditions data. By using segmented acoustic and vibration data, invalid acquisition segments with sampling gaps, time misalignments, and acquisition interruptions are screened out. The retained sound segments, vibration segments, operating condition segments, and the body's abnormal energy conversion benchmark are correlated and recombined to form acoustic and vibration synchronization segments.

3. The method for evaluating the operating status of electrical equipment based on joint sound and vibration detection as described in claim 1, characterized in that, The specific steps for forming the set of acoustic vibration anomalies are as follows: Extract the sound data segment and vibration data segment of the sound-vibration synchronization segment, and perform fluctuation interval identification and abnormal start and end location respectively to obtain the sound abnormal segment, vibration abnormal segment and the corresponding abnormal occurrence time; Based on the abnormal sound segments, abnormal vibration segments and their corresponding occurrence times, the abnormal duration intervals and the corresponding working conditions are matched and organized to obtain the sound and vibration abnormality associated segments. By using the correlation segments of acoustic and vibration anomalies, corresponding acoustic and vibration anomaly segments under the same working condition are recombined to form a set of acoustic and vibration anomaly events.

4. The method for evaluating the operating status of electrical equipment based on joint sound and vibration detection as described in claim 1, characterized in that, The specific steps for generating consistent abnormal source data for acoustic and vibration operating conditions are as follows: Based on the set of acoustic and vibration abnormal events, the acoustic and vibration abnormal source competition resolution model is used to extract the corresponding sound abnormal segment, vibration abnormal segment, abnormal occurrence time, abnormal duration interval and working condition correspondence for each abnormal event. Sound abnormal segments and vibration abnormal segments that belong to the same abnormal occurrence time, the same abnormal duration interval and the same working condition correspondence are paired and organized to obtain acoustic and vibration abnormal pairing data. Based on the sound and vibration anomaly pairing data, focusing on the source types of the target equipment body anomaly, external sound source interference, external vibration impact, structural propagation anomaly, and acquisition anomaly, source pointing mark and correspondence verification are performed on each set of sound and vibration anomaly pairing data to obtain candidate source matching data; Using candidate source matching data, the correspondence between sound abnormal segments, vibration abnormal segments, and operating conditions of each candidate source is collaboratively determined. Candidate sources with matching relationships are determined as consistent sound and vibration operating conditions and retained, while candidate sources with unmatched relationships are determined as inconsistent sound and vibration operating conditions and eliminated, thus generating consistent sound and vibration operating condition abnormal source data.

5. The method for evaluating the operating status of electrical equipment based on joint sound and vibration detection as described in claim 1, characterized in that, The specific construction process of the competition resolution model for acoustic and vibration anomalies is as follows: A model for resolving competition among acoustic and vibration anomaly sources is constructed based on an acoustic and vibration anomaly pairing identification layer, an anomaly source competition discrimination layer, and a working condition consistency screening layer. The sound and vibration anomaly pairing and identification layer, based on the set of sound and vibration anomaly events, organizes the features of sound anomaly segments, vibration anomaly segments, anomaly occurrence time, anomaly duration interval and working condition correspondence, and pairs sound and vibration segments according to the same time interval and the same working condition to generate sound and vibration anomaly pairing feature data. The abnormal source competition discrimination layer, based on the sound and vibration abnormality pairing feature data, performs source pointing marking and competition discrimination according to the source type of the target device body abnormality, external sound source interference, external vibration impact, structural propagation abnormality and acquisition abnormality, and generates candidate source matching data; The working condition consistency screening layer, based on candidate source matching data, collaboratively determines the correspondence between abnormal sound segments, abnormal vibration segments, and working condition correspondence, and retains and merges the candidate sources determined to be consistent sound and vibration working conditions, and outputs consistent sound and vibration working condition abnormal source data.

6. The method for evaluating the operating status of electrical equipment based on joint detection of sound and vibration as described in claim 4, characterized in that, The specific steps for generating consistent abnormal source data for acoustic and vibration operating conditions are as follows: Based on the candidate source matching data, the corresponding sound abnormality segment markers, vibration abnormality segment markers, and working condition corresponding markers are extracted according to the source type of the candidate source to obtain the abnormal occurrence time and abnormal duration interval corresponding to each candidate source. When the sound abnormality segment marker, vibration abnormality segment marker, and operating condition corresponding marker all correspond to the same abnormality occurrence time and the same abnormality duration interval, the candidate source is determined to be a consistent source of sound and vibration operating conditions. When the sound abnormality segment marker, vibration abnormality segment marker, and operating condition corresponding marker do not simultaneously correspond to the same abnormality occurrence time and the same abnormality duration interval, the candidate source is determined to be an inconsistent source of sound and vibration operating conditions. Sources with consistent acoustic and vibration conditions are retained and merged, while sources with inconsistent acoustic and vibration conditions are removed from the candidate source matching data to generate data on anomalous sources with consistent acoustic and vibration conditions.

7. The method for evaluating the operating status of electrical equipment based on joint sound and vibration detection as described in claim 1, characterized in that, The specific steps for generating acoustic vibration energy conservation deviation data are as follows: Based on the consistent abnormal source data of acoustic and vibration working conditions, the corresponding relationship between the sound abnormal segment, vibration abnormal segment and working condition is extracted from each consistent abnormal source of acoustic and vibration working conditions. The energy change within the same abnormal duration interval of the sound abnormal segment and vibration abnormal segment is extracted to obtain the acoustic and vibration energy change data. Based on the acoustic and vibration energy change data, the sound energy change and vibration energy change are recombined according to the consistent abnormal source under the same acoustic and vibration working condition to obtain the corresponding acoustic and vibration energy data. Using the body's abnormal energy conversion benchmark, the deviation comparison of sound energy changes and vibration energy changes in the corresponding sound and vibration energy data is performed to determine whether the corresponding sound and vibration energy data belongs to the body energy consistent state, sound side deviation state, and vibration side deviation state. Based on the determination results, source classification is performed to form sound and vibration energy conservation deviation data.

8. The method for evaluating the operating status of electrical equipment based on joint detection of sound and vibration as described in claim 7, characterized in that, The specific steps for generating acoustic vibration energy conservation deviation data are as follows: Based on the corresponding data of acoustic and vibration energy, extract the sound energy change, vibration energy change and working condition correspondence corresponding to the consistent abnormal source of each acoustic and vibration working condition, and match the body abnormal energy conversion benchmark under the corresponding working condition. When the correspondence between sound energy change and vibration energy change falls within the range corresponding to the abnormal energy conversion benchmark of the body, it is determined to be a state of consistent body energy. When the sound energy change deviates from the abnormal energy conversion benchmark of the body and the vibration energy change does not form a corresponding change, it is determined to be a state of sound side deviation. When the vibration energy change deviates from the abnormal energy conversion benchmark of the body and the sound energy change does not form a corresponding change, it is determined to be a state of vibration side deviation. By combining the vibration side deviation state with the source type label in the candidate source matching data, when the source type label corresponds to external vibration impact, it is classified as external vibration impact associated data; when the source type label corresponds to structural propagation anomaly, it is classified as structural propagation anomaly associated data. The classification results are then organized to obtain acoustic vibration energy conservation deviation data.

9. The method for evaluating the operating status of electrical equipment based on joint detection of sound and vibration as described in claim 1, characterized in that, The specific steps for generating the target electrical equipment's operational status evaluation results are as follows: Based on the acoustic and vibration energy conservation deviation data, the source type, energy deviation status and working condition correspondence in the acoustic and vibration working condition consistency anomaly source data are extracted. Ontology association verification and source marking are performed on each anomaly source to obtain source correction anomaly data. Based on the source correction of abnormal data, the data corresponding to external sound source interference, external vibration impact, structural propagation anomalies and acquisition anomalies are merged and organized to obtain non-body anomaly records. The anomaly sources that meet the body anomaly energy conversion benchmark are merged and organized to obtain body anomaly evaluation records. By utilizing non-physical anomaly records and physical anomaly evaluation records, the anomaly sources, acoustic and vibration deviation states, and corresponding operating conditions of the target electrical equipment are reorganized to generate the operating status evaluation results of the target electrical equipment.

10. The method for evaluating the operating status of electrical equipment based on joint sound and vibration detection as described in claim 9, characterized in that, The specific steps for obtaining non-ontology anomaly records and ontology anomaly evaluation records are as follows: Based on the source correction of abnormal data, extract the source type, sound abnormal segment, vibration abnormal segment, working condition correspondence and sound and vibration energy conservation deviation data corresponding to each abnormal source, and classify the data corresponding to external sound source interference, external vibration impact, structural propagation abnormality and acquisition abnormality into non-body abnormality categories according to the source type. Data classified as non-entity anomalies are merged according to the time of occurrence of the anomaly, the duration of the anomaly, and the correspondence with the working conditions. Duplicate sound anomaly segments and vibration anomaly segments are deleted to obtain non-entity anomaly records. For anomaly sources that meet the energy conversion benchmark of the device body, the data on the target device body anomaly, anomaly occurrence time, anomaly duration interval, and acoustic and vibration energy conservation deviation are merged, and the corresponding sound anomaly segments, vibration anomaly segments, and working condition correspondence are associated and saved to obtain the body anomaly evaluation record.

Images