Fan noise detection monitoring method and system, computer equipment and storage medium

Abstract

The invention relates to the technical field of noise detection, in particular to a fan noise detecting and monitoring method and system, computer equipment and a storage medium, and the fan noise detecting and monitoring method comprises the following steps: collecting audio data during operation of a fan based on an acoustic sensor, performing frequency decomposition on an acoustic signal through wavelet transform, and obtaining an acoustic signal; calculating the energy spectrum density under the difference frequency, identifying the distribution characteristics of the multi-band energy, and carrying out the preliminary classification of the noise mode generated by the fan. According to the method, wavelet transformation is carried out on the audio data during operation of the fan, the analysis capability of noise data is improved, more detailed frequency decomposition is allowed to be carried out on sound wave signals, the recognition precision of the noise mode of the fan is enhanced, continuous operation monitoring and state evaluation of the fan are improved through a real-time monitoring and automatic recognition mechanism, and the recognition efficiency of the fan is improved. The threshold parameter of noise monitoring is dynamically adjusted, and the change of environment and the change of equipment conditions can be immediately responded, so that the accuracy and response speed of event detection are optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of noise detection, and in particular to a method, system, computer equipment and storage medium for detecting and monitoring fan noise. Background Art

[0002] Noise detection technology is a branch of acoustic engineering, which mainly involves the measurement and analysis of environmental, industrial and other types of noise. It uses various sensors and measuring equipment, such as sound level meters and spectrum analyzers, to capture sound data and analyze these data through signal processing technology to determine the level, frequency and source of the sound. The goal is to evaluate the impact of noise on health, work efficiency and quality of life.

[0003] Among them, the fan noise detection and monitoring method involves the use of specific technologies to monitor and evaluate the noise generated by the fan during operation to ensure the normal operation of the fan and reduce the impact of noise on the surrounding environment, including the use of acoustic sensors to continuously monitor the sound generated by the fan and identify the noise source and noise level through real-time data analysis, which helps to detect mechanical failures and performance problems in a timely manner, so as to carry out repairs or adjustments to prevent efficiency loss and extend equipment life.

[0004] Existing technologies lack immediacy in data processing and are unable to achieve dynamic identification and instant classification of noise events, which limits the accurate assessment and timely intervention of wind turbine status. They lack technologies with real-time monitoring and dynamic threshold adjustment functions and cannot adapt to changes in equipment and environmental conditions, leading to misjudgment or omission of noise levels. For example, equipment aging or increased environmental noise is not reflected in the monitoring data in a timely manner, which will lead to incorrect maintenance decisions, increasing maintenance costs and ignoring actual problems that need to be solved, thereby affecting the overall operating efficiency and life of the equipment. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a fan noise detection and monitoring method, system, computer equipment and storage medium.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a method for detecting and monitoring fan noise, comprising the following steps: S1: Based on the acoustic sensor, the audio data of the fan during operation is collected, the sound wave signal is decomposed by wavelet transform, the energy spectrum density under the difference frequency is calculated, the distribution characteristics of multi-band energy are identified, and the noise mode generated by the fan is preliminarily classified to form a frequency feature set; S2: Analyze the noise characteristics of the gearbox and the motor in the fan according to the frequency feature set, automatically identify the noise patterns of the components, and obtain a component noise identification set; S3: Based on the component noise identification set, real-time noise monitoring is performed, the fan noise pattern is instantly identified and classified, and the current operating state of the fan is evaluated to obtain an operating state evaluation log; S4: Analyze the operation status evaluation log, determine the frequency and duration of noise events through change point analysis, adjust the threshold parameters of noise monitoring, optimize the accuracy and response speed of event detection, and obtain the threshold adjustment index; S5: Based on the threshold adjustment index, continuously monitor the fan noise, identify and record the noise events exceeding the standard by comparing the real-time data with the threshold, and determine the potential fault type of the fan to obtain the fault identification record; S6: Based on the fault identification record, adjust or repair the fan, eliminate the noise source, re-monitor the noise level of the fan, verify whether the fan returns to normal operation, and continuously monitor the noise control effect to obtain the maintenance effect verification result.

[0007] As a further solution of the present invention, the frequency feature set includes noise source location, frequency range, and energy level; the component noise identification set includes sound spectrum, noise level, and abnormal frequency identification results; the operating status evaluation log includes equipment status code, operating time record, and performance score; the threshold adjustment index includes sensitivity level, response speed range, and detection accuracy index; the fault identification record includes fault timestamp, fault impact degree, and fault recovery time; the maintenance effect verification result includes noise reduction ratio, operating smoothness index, and post-maintenance frequency.

[0008] As a further solution of the present invention, the steps of analyzing the noise characteristics of the gearbox and the motor in the fan according to the frequency feature set, automatically identifying the noise mode of the components, and obtaining the component noise identification set are specifically as follows: S201: selecting the noise frequency bands of the gearbox and the motor target according to the frequency feature set, analyzing the energy and time series of the frequency bands, comparing the differences between the data and the known noise patterns, and obtaining the noise difference analysis results; S202: Analyze the noise difference analysis results, compare and identify component noise patterns, record successfully matched data, and generate a component noise comparison table; S203: Using the component noise comparison table, setting a monitoring threshold, implementing real-time monitoring, and identifying and classifying noise patterns to obtain a component noise identification set.

[0009] As a further solution of the present invention, based on the component noise identification set, noise monitoring is performed in real time, the fan noise pattern is instantly identified and classified, and the current operating state of the fan is evaluated. The steps of obtaining the operating state evaluation log are specifically as follows: S301: Based on the component noise identification set, the acoustic wave signal is captured in real time, the noise of the difference frequency is separated, and the separated noise is matched with a preset pattern, the noise pattern is identified, and a noise matching result is generated; S302: Based on the noise matching result, perform data comparison and analyze data differences, identify component behaviors that deviate from normal operating states, evaluate and record the current operating state of the wind turbine, and obtain deviation behavior records; S303: Utilize the deviation behavior record to analyze the change trend in the noise data, adjust the monitoring threshold parameters, and match the real-time monitoring requirements to obtain an operation status evaluation log.

[0010] As a further solution of the present invention, the data comparison and analysis of data differences are performed to identify component behaviors that deviate from normal operating conditions, according to the formula:

[0011] Get the deviant behavior record, where, represents the deviation value of each data point extracted from the noise matching result, is the average of the deviation values, is the total number of deviation values, is the root mean square of the deviation.

[0012] As a further solution of the present invention, the operation status evaluation log is analyzed, and the frequency and duration of noise events are determined through change point analysis, and the threshold parameters of noise monitoring are adjusted to optimize the accuracy and response speed of event detection. The steps of obtaining the threshold adjustment index are specifically as follows: S401: Analyze the operation status evaluation log, determine the frequency and duration of noise events through change point analysis, and integrate key noise event features to obtain noise event data; S402: Based on the noise event data, analyze the monitoring threshold adjustment requirements, implement monitoring parameter optimization, match the changes in real-time data, and update the monitoring settings to obtain monitoring parameter configuration; S403: Execute the monitoring parameter configuration, test the effect of the new parameter through real-time data monitoring, verify whether the implementation of the threshold adjustment meets the standard, and obtain the threshold adjustment index.

[0013] As a further solution of the present invention, based on the threshold adjustment index, the fan noise is continuously monitored, and by comparing the real-time data with the threshold, the noise events exceeding the standard are identified and recorded, and the potential fault type of the fan is determined. The steps of obtaining the fault identification record are specifically as follows: S501: Based on the threshold adjustment index, continuously monitor the fan noise, capture the real-time sound wave signal, compare it with the threshold, automatically mark the signal exceeding the threshold, and obtain the exceeding signal record; S502: Based on the excessive signal record, data clustering processing is performed on the marked events, and the potential fault type of the associated wind turbine component is determined by analyzing the characteristic frequency and intensity of the event to obtain a fault identification record.

[0014] A fan noise detection and monitoring system, the system comprising: The sound feature extraction module collects audio data from the fan when it is running based on the acoustic sensor, applies wavelet transform to perform frequency decomposition on the signal, calculates the energy spectrum density under the difference frequency, analyzes the multi-band energy distribution characteristics, classifies the noise mode, and obtains the frequency feature set; The noise identification module analyzes the noise characteristics of the gearbox and the motor in the fan based on the frequency feature set, automatically identifies the component noise mode, and obtains the component noise identification set; The state monitoring module monitors the fan noise in real time based on the component noise identification set, identifies and classifies the noise pattern, and evaluates the current operating state of the fan to obtain an operating state evaluation log; The event adjustment module analyzes the operation status evaluation log, determines the frequency and duration of noise events through change point analysis, adjusts the noise monitoring threshold parameters, and optimizes the detection accuracy and response speed to obtain the threshold adjustment index; Based on the threshold adjustment index, the fault processing module continuously monitors the fan noise, compares the real-time data with the threshold, identifies and records the noise events that exceed the standard, analyzes the potential fault types of the fan, adjusts or repairs the fan, and re-monitors to determine whether the fan has returned to normal operation, thereby obtaining the maintenance effect verification result.

[0015] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the fan noise detection and monitoring system as described above when executing the computer program.

[0016] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the fan noise detection and monitoring method as described above.

[0017] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by adopting wavelet transform to the audio data of the fan during operation, the parsing capability of noise data is improved, allowing more detailed frequency decomposition of the sound wave signal, and enhancing the recognition accuracy of the fan noise pattern. Through real-time monitoring and automatic recognition mechanism, the continuous operation monitoring and status evaluation of the fan are improved, and the threshold parameters of noise monitoring are dynamically adjusted, which can respond to environmental changes and changes in equipment conditions in real time, thereby optimizing the accuracy and response speed of event detection, effectively reducing the risk of operation interruption due to faults or performance problems, and improving the overall efficiency and safety of the fan. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the workflow of the present invention; Figure 2 This is a detailed flow chart of S1 of the present invention; Figure 3 This is a detailed flow chart of S2 of the present invention; Figure 4 This is a detailed flow chart of S3 of the present invention; Figure 5 This is a detailed flow chart of S4 of the present invention; Figure 6 This is a detailed flow chart of S5 of the present invention; Figure 7 This is a detailed flow chart of S6 of the present invention; Figure 8 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0021] Embodiment 1 See also Figure 1 The present invention provides a technical solution: a method for detecting and monitoring fan noise, comprising the following steps: S1: Based on the acoustic sensor, the audio data of the fan during operation is collected, the sound wave signal is decomposed by wavelet transform, the energy spectrum density under the difference frequency is calculated, the distribution characteristics of multi-band energy are identified, and the noise mode generated by the fan is preliminarily classified to form a frequency feature set; S2: Based on the frequency feature set, analyze the noise characteristics of the gearbox and motor in the fan, automatically identify the noise mode of the components, and obtain the component noise identification set; S3: Based on the component noise identification set, real-time noise monitoring is performed to instantly identify and classify the fan noise pattern, and the current operating status of the fan is evaluated to obtain the operating status evaluation log; S4: Analyze the operation status evaluation log, determine the frequency and duration of noise events through change point analysis, and adjust the threshold parameters of noise monitoring to optimize the accuracy and response speed of event detection and obtain the threshold adjustment index; S5: Based on the threshold adjustment index, the fan noise is continuously monitored, and by comparing the real-time data with the threshold, the noise events exceeding the standard are identified and recorded, and the potential fault type of the fan is determined to obtain the fault identification record; S6: Based on the fault identification record, adjust or repair the fan, eliminate the noise source, re-monitor the noise level of the fan, verify whether the fan has returned to normal operation, and continuously monitor the noise control effect to obtain the maintenance effect verification result.

[0022] The frequency feature set includes noise source location, frequency range, and energy level. The component noise identification set includes sound spectrum, noise level, and abnormal frequency identification results. The operation status assessment log includes equipment status code, operation time record, and performance score. The threshold adjustment indicators include sensitivity level, response speed range, and detection accuracy index. The fault identification record includes fault timestamp, fault impact degree, and fault recovery time. The maintenance effect verification results include noise reduction ratio, operation smoothness index, and post-maintenance frequency.

[0023] See also Figure 2 , based on the acoustic sensor to collect the audio data of the fan when it is running, the sound wave signal is decomposed by wavelet transform, the energy spectrum density under the difference frequency is calculated, the distribution characteristics of multi-band energy are identified, and the noise mode generated by the fan is preliminarily classified. The steps to form the frequency feature set are as follows: S101: collecting audio data when the fan is running based on an acoustic sensor, applying wavelet transform to perform frequency decomposition on the audio data, measuring the energy spectrum density of each frequency band from the decomposed data, performing statistical analysis on the measured values, and obtaining frequency spectrum analysis records; Based on the acoustic sensor, the audio data of the fan during operation is collected, the original waveform of the signal is extracted, and the audio signal is segmented and intercepted with each segment having a fixed sampling period. The wavelet transform is used to decompose the audio signal into multi-scale frequency components, and the power spectral density of each component is extracted according to the frequency range. The power value of each frequency band is measured and normalized, and the energy distribution statistics are calculated. Further, an energy frequency band matrix is ​​created according to the statistical value, and the power values ​​of different frequency bands are divided and filled into the matrix according to the interval. After the statistical matrix is ​​completed, the results are summarized and the frequency band data records are generated.

[0024] S102: Analyze the frequency spectrum analysis record, identify the frequency bands with high and low energy according to the energy distribution data, and classify the noise mode according to the frequency bands, record each classification result, and obtain a frequency feature set; Analyze the frequency spectrum analysis records, extract the data groups in the frequency band matrix, sort the energy data of each frequency band, mark the frequency bands with the highest and lowest energy according to the sorting results, and classify them into high-energy bands and low-energy bands based on the data distribution pattern of the frequency band matrix. Compare the classified frequency bands with the predefined noise patterns, number the noise classifications according to the pattern matching results and record the classification numbers, and summarize the relevant parameters of the classified frequency bands to form a noise pattern feature table.

[0025] See also Figure 3 , according to the frequency feature set, analyze the noise characteristics of the gearbox and motor in the fan, automatically identify the noise mode of the components, and obtain the steps of the component noise identification set as follows: S201: selecting the noise frequency bands of the gearbox and motor targets according to the frequency feature set, analyzing the energy and time series of the frequency bands, comparing the differences between the data and the known noise patterns, and obtaining the noise difference analysis results; According to the frequency feature set, the target noise frequency band of the gearbox and motor is selected, the energy value and time series of the specific frequency band are extracted from the frequency feature set, and the data is divided into multiple equal-length segments by time series segmentation. The energy values ​​in each segment are summarized, and an energy time table for each segment is generated. The high-energy peak and low-energy area in the energy time table are compared, and the features are compared with the parameters in the known noise pattern database. Key indicators such as energy difference, duration, and frequency band distribution are used to determine whether there is a significant deviation. The specific values ​​and comparison results of the difference features are recorded, and the noise difference analysis results are generated after data matching is completed.

[0026] S202: Analyze the noise difference analysis results, compare and identify component noise patterns, record successfully matched data, and generate a component noise comparison table; The noise difference feature set is analyzed to extract feature items that match the known noise patterns. It is checked whether the matching features meet specific threshold conditions. The features that meet the threshold are classified and recorded. According to the frequency band to which the matched noise features belong, the corresponding mechanical components are marked and associated with the existing noise patterns. A matching list between the components and the noise patterns is generated. At the same time, feature items that do not meet the matching criteria are eliminated. The generated matching list is optimized and formatted to form a comprehensive component noise correspondence file, and finally a component noise pattern list is obtained.

[0027] S203: using the component noise comparison table, setting a monitoring threshold, implementing real-time monitoring, and identifying and classifying noise patterns to obtain a component noise identification set; Utilize the component noise pattern list to extract characteristic data of high-frequency matching noise patterns, set monitoring thresholds, and classify noise signals into normal and abnormal modes based on the comparison threshold range of the acoustic wave signals captured in real time. Classify and identify the signal characteristics of each mode, and summarize the signal characteristic results by classification. Further classify the signals according to frequency bands and energy ranges, record the properties of each type of noise signal and the name of the component to which it belongs, generate structured files for all identified and classified noise patterns, and obtain the component noise identification set after completing monitoring and classification.

[0028] See also Figure 4 Based on the component noise identification set, real-time noise monitoring is performed to instantly identify and classify the fan noise pattern and evaluate the current operating status of the fan. The specific steps for obtaining the operating status evaluation log are as follows: S301: Based on the component noise recognition set, the acoustic wave signal is captured in real time, the noise of the difference frequency is separated, and the separated noise is matched with the preset pattern, the noise pattern is identified, and the noise matching result is generated; Based on the component noise identification set, the acoustic wave signal captured in real time is preprocessed and divided into continuous samples within a fixed time window. Each sample is extracted through the frequency separation module to extract signals of different frequency components. The components are sorted according to their energy values ​​and the signal amplitudes of the difference frequencies are recorded. The corresponding energy spectral density is calculated for each frequency component, and the energy spectral density is matched with the preset noise pattern reference data. The deviation value of each signal component is calculated through matching difference analysis, and the signal components that meet the noise pattern conditions are screened out. The signal components that meet the conditions are numbered and classified to generate noise matching results.

[0029] S302: Based on the noise matching result, perform data comparison and analyze data differences, identify component behaviors that deviate from normal operating states, evaluate and record the current operating state of the wind turbine, and obtain deviation behavior records; Compare data and analyze data differences to identify component behaviors that deviate from normal operating conditions, according to the formula: ; Get the deviant behavior record, where, represents the deviation value of each data point extracted from the noise matching result, is the average of the deviation values, is the total number of deviation values, is the root mean square of the deviation, a measure of how much the component behavior deviates from normal operating conditions; The deviation value extracted from the noise matching result is ,but ,calculate , substituting the value into the formula we get: ; ; The results show that the average deviation is 0.158, indicating that the behavior of the component has deviated significantly from the normal state and requires further inspection and repair.

[0030] S303: Analyze the change trend in the noise data by using the deviation behavior record, adjust the monitoring threshold parameters, and match the real-time monitoring requirements to obtain the operation status evaluation log; Utilize the deviation behavior records to screen out the noise signals with high-frequency anomalies, plot the data trend of the signals as a time series curve, calculate the noise change rate within a specific time period based on the curve change trend, further analyze the data whose change rate exceeds the monitoring threshold, redefine the monitoring threshold parameter range based on the analysis results, set the new parameter value to real-time monitoring, adjust the monitoring sensitivity and range, ensure that the monitoring system is consistent with the real-time monitoring requirements, classify and record the signals captured by the monitoring after the parameter adjustment, and generate an operation status evaluation log.

[0031] See also Figure 5 , analyze the operation status evaluation log, determine the frequency and duration of noise events through change point analysis, and adjust the threshold parameters of noise monitoring to optimize the accuracy and response speed of event detection. The specific steps to obtain the threshold adjustment index are as follows: S401: Analyze the operation status assessment log, determine the frequency and duration of noise events through change point analysis, and integrate key noise event features to obtain noise event data; Analyze the operation status assessment log, extract the noise data recorded in each time period from the log, classify the noise data according to the timeline of the event, mark the start time and end time of each event, calculate the time interval and record the duration of the event, extract the frequency band data corresponding to the event according to the frequency, form a matrix table of frequency components and event duration, integrate the key frequency band data to form an event feature set, locate the points with large frequency changes through change point analysis and associate them with specific noise events, summarize the data of key noise events, and generate noise event data.

[0032] S402: Based on the noise event data, analyze the monitoring threshold adjustment requirements, implement monitoring parameter optimization, match the changes in real-time data, and update the monitoring settings to obtain monitoring parameter configuration; Based on the noise event data, relevant parameters of key characteristic frequency bands and duration are extracted, the range of frequency distribution and the change amplitude of energy density in each frequency band are measured, and the adaptability of the current monitoring threshold is analyzed according to the data. New monitoring range and sensitivity requirements are defined, and incompatible data points are readjusted. The response time and frequency resolution of real-time monitoring are set by updating the monitoring threshold parameters. The new parameter configuration is applied to the monitoring system and tested. The real-time noise data captured in the test is compared with the updated threshold in real time to verify the accuracy of parameter optimization and confirm its adaptability, and obtain the monitoring parameter configuration.

[0033] S403: Execute monitoring parameter configuration, test the effect of new parameters through real-time data monitoring, verify whether the implementation of threshold adjustment meets the standard, and obtain threshold adjustment indicators; Execute monitoring parameter configuration, load optimized parameters, perform frequency band separation on the noise signal captured in real time, screen signal components that meet the monitoring threshold conditions, calculate the energy density and duration of the signal components, and compare the results with the set performance standards by judging the matching status of each type of signal. According to the comparison results, correct the parameter configuration, continuously verify the corrected parameters in multiple noise data scenarios, record the key performance data during the verification process, and obtain the threshold adjustment index.

[0034] See also Figure 6 Based on the threshold adjustment index, the fan noise is continuously monitored. By comparing the real-time data with the threshold, the noise events exceeding the standard are identified and recorded, and the potential fault type of the fan is determined. The specific steps for obtaining the fault identification record are as follows: S501: Based on the threshold adjustment index, the fan noise is continuously monitored, the real-time sound wave signal is captured, and compared with the threshold, and the signal exceeding the threshold is automatically marked to obtain the exceeding signal record; The sensor is used to capture the acoustic wave signals generated during the operation of the fan in real time, and the signals are segmented and stored as sample data of a fixed time length. The spectrum of each segment of data is decomposed to extract the frequency components. The components are screened according to the preset monitoring threshold range, and the amplitude of each component is compared with the threshold range. The signal with an amplitude exceeding the threshold is recorded as an abnormal noise event. The timestamp, frequency, and energy value of each abnormal event are recorded in the abnormal event log. The events are arranged in chronological order and the source signal category is marked to generate a record of excessive signals.

[0035] S502: Based on the excessive signal record, data clustering processing is performed on the marked events, and the potential fault types of the associated wind turbine components are determined by analyzing the characteristic frequency and intensity of the events to obtain a fault identification record; Analyze the characteristic frequency and intensity of the event, and cluster the event characteristics according to the formula: ; Calculate the internal variance of each cluster , as a measure of clustering quality, where is the total number of clusters, It is The feature vector of each event, including frequency and intensity, It is The center vector of each cluster, Representative Clusters.

[0036] Total number of clusters Set to 3, event feature vector Includes the frequency and intensity extracted from the sound signal, for example, the frequency is 150Hz and the intensity is 20dB; the cluster center vector Represents the average feature vector of each cluster, which is calculated iteratively according to the clustering algorithm. For example, for the first cluster, the frequency is 100Hz and the intensity is 15dB; the internal variance It is measured by calculating the sum of the squares of the distances between each vector and its cluster center, which is used to evaluate the tightness of the clustering. Considering the clustering example, let , and set : right : ; right : ; right : ; but Calculated as: ; The result shows that the internal variance is large, indicating that the differences within the clusters are large, and it is necessary to adjust the clustering parameters or increase the number of clusters to achieve better classification results.

[0037] See also Figure 7 Based on the fault identification record, adjust or repair the fan, eliminate the noise source, re-monitor the noise level of the fan, verify whether the fan has returned to normal operation, and continuously monitor the noise control effect. The specific steps to obtain the maintenance effect verification result are as follows: S601: Based on the fault identification record, perform a fan adjustment or repair operation, replace or repair the identified problematic component, and recalibrate the fan operation settings to obtain a fan update status; Based on the fault identification records, determine the fan components that need to be adjusted or repaired, disassemble the target components identified as faulty for inspection, measure the key parameters of the components such as size, tolerance and wear degree, compare the measured values ​​with the standard parameters of the components, repair or replace the components that deviate from the standard range, reinstall the repaired components and check the connection with surrounding components, adjust the position of the assembled components to ensure balance and stability during operation, recalibrate the fan operation settings after completing the component adjustment, adjust the motor power, speed and gearbox ratio, record all repair and calibration parameters, and obtain the fan update status.

[0038] S602: based on the updated status of the fan, restart the adjusted fan, continuously monitor its operating status, and compare the noise data before and after the repair to determine whether the noise level reaches the normal range, and obtain the noise level assessment result; Capture the acoustic wave signals during operation in real time, continuously record the energy distribution and frequency changes of the acoustic wave signals, compare the recorded noise data with the data before repair, compare the frequency range, amplitude and time series changes of the data, measure the difference between the frequency bands before and after repair, calculate the noise amplitude reduction rate of each frequency band, analyze the changes in energy distribution trends by time period, confirm whether the noise level after repair is within the preset range based on the analysis results, mark and record the abnormal points that do not meet the range, and generate the noise level assessment results.

[0039] S603: Based on the noise level assessment results, continuously evaluate the noise control effect of the fan, verify the effectiveness of maintenance through regular performance evaluation, and obtain the maintenance effect verification results; Based on the noise level assessment results, extract the noise control effect data in the records, regularly analyze the changing trends in the monitoring data, compare the noise frequency and energy distribution in different time periods, identify the rules of noise level recovery, match the changing rules with the original noise level, confirm the stability of noise control, check whether there are new abnormal changes in the monitoring data, record and classify the existing potential problems item by item, and obtain the maintenance effect verification results.

[0040] See also Figure 8 , fan noise detection and monitoring system, the system includes: The sound feature extraction module collects audio data from the fan when it is running based on the acoustic sensor, applies wavelet transform to perform frequency decomposition on the signal, calculates the energy spectrum density under the difference frequency, analyzes the multi-band energy distribution characteristics, classifies the noise mode, and obtains the frequency feature set; The noise identification module analyzes the noise characteristics of the gearbox and motor in the fan based on the frequency feature set, automatically identifies the component noise pattern, and obtains the component noise identification set; The status monitoring module monitors the fan noise in real time based on the component noise identification set, identifies and classifies the noise pattern, and evaluates the current operating status of the fan to obtain the operating status evaluation log; The event adjustment module analyzes the operation status evaluation log, determines the frequency and duration of noise events through change point analysis, adjusts the noise monitoring threshold parameters, and optimizes the detection accuracy and response speed to obtain the threshold adjustment index; The fault handling module continuously monitors the fan noise based on the threshold adjustment index, compares the real-time data with the threshold, identifies and records the noise events that exceed the standard, analyzes the potential fault types of the fan, adjusts or repairs the fan, and re-monitors to determine whether the fan has returned to normal operation and obtain the maintenance effect verification result.

[0041] A computer device comprises a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the fan noise detection and monitoring system as described above is implemented.

[0042] A computer-readable storage medium stores a computer program, which implements the steps of the above-mentioned fan noise detection and monitoring method when executed by a processor.

[0043] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A fan noise detection and monitoring method, characterized in that: The following steps are involved: The acoustic sensor collects audio data when the fan is running, decomposes the sound wave signal by wavelet transform, calculates the energy spectrum density at the difference frequency, identifies the distribution characteristics of multi-band energy, and preliminarily classifies the noise pattern generated by the fan to form a frequency feature set. According to the frequency feature set, the noise characteristics of the gearbox and the motor in the fan are analyzed, the noise patterns of the components are automatically identified, and a component noise identification set is obtained; Based on the component noise identification set, real-time noise monitoring is performed, the fan noise pattern is instantly identified and classified, and the current operating state of the fan is evaluated to obtain an operating state evaluation log; Analyze the operation status evaluation log, determine the frequency and duration of noise events through change point analysis, adjust the threshold parameters of noise monitoring, optimize the accuracy and response speed of event detection, and obtain the threshold adjustment index; Based on the threshold adjustment index, the fan noise is continuously monitored, and by comparing the real-time data with the threshold, the noise events exceeding the standard are identified and recorded, and the potential fault type of the fan is determined to obtain the fault identification record; Based on the fault identification record, the fan is adjusted or repaired, the noise source is eliminated, the noise level of the fan is re-monitored, it is verified whether the fan has returned to a normal operating state, and the noise control effect is continuously monitored to obtain a maintenance effect verification result.

2. The fan noise detection and monitoring method according to claim 1, characterized in that: The frequency feature set includes noise source location, frequency range, and energy level; the component noise identification set includes sound spectrum, noise level, and abnormal frequency identification results; the operation status assessment log includes equipment status code, operation time record, and performance score; the threshold adjustment index includes sensitivity level, response speed range, and detection accuracy index; the fault identification record includes fault timestamp, fault impact degree, and fault recovery time; the maintenance effect verification result includes noise reduction ratio, operation stability index, and post-maintenance frequency.

3. The fan noise detection and monitoring method according to claim 1, characterized in that: According to the frequency feature set, the noise characteristics of the gearbox and the motor in the fan are analyzed, the noise patterns of the components are automatically identified, and the steps of obtaining the component noise identification set are specifically as follows: According to the frequency feature set, the noise frequency bands of the gearbox and the motor targets are selected, the energy and time series of the frequency bands are analyzed, the differences between the data and the known noise patterns are compared, and the noise difference analysis results are obtained; Analyze the noise difference analysis results, compare and identify component noise patterns, record successfully matched data, and generate a component noise comparison table; The component noise comparison table is used to set a monitoring threshold, implement real-time monitoring, and identify and classify noise patterns to obtain a component noise identification set.

4. The fan noise detection and monitoring method according to claim 1, characterized in that: Based on the component noise identification set, real-time noise monitoring is performed, the fan noise pattern is instantly identified and classified, and the current operating state of the fan is evaluated. The specific steps of obtaining the operating state evaluation log are as follows: Based on the component noise identification set, the acoustic wave signal is captured in real time, the noise of the difference frequency is separated, and the separated noise is matched with the preset pattern, the noise pattern is identified, and the noise matching result is generated; Based on the noise matching result, data comparison and data difference analysis are performed to identify component behaviors that deviate from normal operating states, evaluate and record the current operating state of the fan, and obtain deviation behavior records; The deviation behavior records are used to analyze the change trend in the noise data, adjust the monitoring threshold parameters, and match the real-time monitoring requirements to obtain the operation status evaluation log.

5. The fan noise detection and monitoring method according to claim 4, characterized in that: Perform the data comparison and analyze the data differences to identify component behaviors that deviate from normal operating conditions, according to the formula: ； Get the deviant behavior record, where, represents the deviation value of each data point extracted from the noise matching result, is the average of the deviation values, is the total number of deviation values, is the root mean square of the deviation.

6. The fan noise detection and monitoring method according to claim 1, characterized in that: The operation status evaluation log is analyzed, and the frequency and duration of noise events are determined through change point analysis, and the threshold parameters of noise monitoring are adjusted to optimize the accuracy and response speed of event detection. The steps of obtaining the threshold adjustment index are specifically as follows: Analyze the operation status assessment log, determine the frequency and duration of noise events through change point analysis, and integrate key noise event features to obtain noise event data; Based on the noise event data, analyze the monitoring threshold adjustment requirements, implement monitoring parameter optimization, match the changes in real-time data, and update the monitoring settings to obtain monitoring parameter configuration; Execute the monitoring parameter configuration, test the effect of the new parameters through real-time data monitoring, verify whether the implementation of the threshold adjustment meets the standards, and obtain the threshold adjustment index.

7. The fan noise detection and monitoring method according to claim 1, characterized in that: Based on the threshold adjustment index, the fan noise is continuously monitored, and by comparing the real-time data with the threshold, the noise events exceeding the standard are identified and recorded, and the potential fault type of the fan is determined. The steps of obtaining the fault identification record are specifically as follows: Based on the threshold adjustment index, the fan noise is continuously monitored, the real-time sound wave signal is captured, and compared with the threshold, and the signal exceeding the threshold is automatically marked to obtain the exceeding signal record; Based on the excessive signal record, data clustering processing is performed on the marked events, and the potential fault types of the associated wind turbine components are determined by analyzing the characteristic frequency and intensity of the events to obtain a fault identification record.

8. The fan noise detection and monitoring system is characterized by: The fan noise detection and monitoring system is used to execute the fan noise detection and monitoring method according to any one of claims 1 to 7, and the system comprises: The sound feature extraction module collects audio data from the fan when it is running based on the acoustic sensor, applies wavelet transform to perform frequency decomposition on the signal, calculates the energy spectrum density under the difference frequency, analyzes the multi-band energy distribution characteristics, classifies the noise mode, and obtains the frequency feature set; The noise identification module analyzes the noise characteristics of the gearbox and the motor in the fan based on the frequency feature set, automatically identifies the component noise mode, and obtains the component noise identification set; The state monitoring module monitors the fan noise in real time based on the component noise identification set, identifies and classifies the noise pattern, and evaluates the current operating state of the fan to obtain an operating state evaluation log; The event adjustment module analyzes the operation status evaluation log, determines the frequency and duration of noise events through change point analysis, adjusts the noise monitoring threshold parameters, and optimizes the detection accuracy and response speed to obtain the threshold adjustment index; Based on the threshold adjustment index, the fault processing module continuously monitors the fan noise, compares the real-time data with the threshold, identifies and records the noise events that exceed the standard, analyzes the potential fault types of the fan, adjusts or repairs the fan, and re-monitors to determine whether the fan has returned to normal operation, thereby obtaining the maintenance effect verification result.

9. A computer device comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor implements the fan noise detection and monitoring system described in claim 8 when executing the computer program.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the fan noise detection and monitoring method according to any one of claims 1 to 7 are implemented.

Images