Equipment health optimal monitoring point selection method and system

Abstract

The invention relates to an equipment health optimal monitoring point selection method and system, and relates to the field of equipment monitoring, and the method comprises the steps: collecting sound signals of monitoring points around monitored equipment; determining a signal spectrum and a signal attenuation based on the sound signal; determining a sound source position according to the signal spectrum and the signal attenuation; calculating a signal-to-noise ratio based on the signal spectrum, matching and determining a noise reduction level, and weighting according to the noise reduction level and the sound source position to obtain a comprehensive score; calculating a comprehensive score in combination with the sound source position and the noise reduction level; according to the monitoring points and the sound signals, calculating a fluctuation ratio, and screening out a position with the fluctuation ratio smaller than a stable threshold value and higher than a score threshold value in the comprehensive score as a secondary monitoring point; determining the position distribution of secondary monitoring points according to the sound source position and the equipment structure; and selecting a point location with the maximum coverage degree from the position distribution, and determining the point location with the highest coincidence degree as the optimal point location based on matching of the signal spectrum and the fault spectrum. The application has the effects of improving the monitoring effect and realizing high-precision monitoring.

Description

Technical Field

[0001] The present invention relates to the field of equipment monitoring, and in particular to a method and system for selecting optimal monitoring points for equipment health. Background Art

[0002] In industrial production, production equipment is prone to malfunction during operation, so the inspection of industrial equipment is the key to ensuring the stable operation of the equipment.

[0003] When a fault occurs, the sound waves of the equipment will change, generating abnormal noise and changes in the frequency of abnormal sounds. Currently, the equipment status can be judged based on the sound differences through manual naked ear monitoring or vibration noise analyzers, providing a convenient way to monitor and repair equipment faults, helping to promptly discover potential equipment problems and improve production efficiency and safety.

[0004] In actual use, manual listening is highly subjective, resulting in an increase in the rate of product failures and the inability to achieve high-precision monitoring. For early faults such as slight bearing wear and loose parts, the sound waves generated by the analyzer may exhibit nonlinear characteristics, and such weak features cannot be effectively extracted, resulting in poor monitoring results and the inability to achieve high-precision monitoring. Summary of the Invention

[0005] In order to improve the monitoring effect and achieve high-precision monitoring, the present invention provides a method and system for selecting optimal monitoring points for equipment health.

[0006] In a first aspect, the present invention provides a method for selecting optimal monitoring points for equipment health, which adopts the following technical solution: A method for selecting optimal monitoring points for equipment health, comprising: Collect sound signals from preset monitoring points around the monitored equipment; determining a signal spectrum and a signal attenuation amount based on the sound signal; determining a sound source position according to the signal spectrum and the signal attenuation; Based on the signal spectrum, the signal-to-noise ratio of the preset monitoring point signal is calculated, the noise reduction level of each position is matched and determined, and a comprehensive score is obtained according to the weighting of the noise reduction level and the sound source position; Calculating a comprehensive score of the monitoring point based on the sound source location and the noise reduction level; Calculating the fluctuation rates under various working conditions based on the monitoring points and the corresponding sound signals, and selecting locations where the fluctuation rates are less than a preset stability threshold and the comprehensive scores are higher than a preset score threshold as secondary monitoring points; Determining the position distribution of the secondary monitoring points according to the sound source position and the preset equipment structure; A point with the largest coverage is selected from the position distribution, and a point with the highest overlap is determined as the optimal point based on matching between the signal spectrum and a preset fault spectrum.

[0007] By adopting the above technical solution, sound signals are collected and the signal spectrum and attenuation are extracted. The signal-to-noise ratio and noise reduction level are calculated based on the spectrum. Then, the sound source location, operating condition fluctuation rate, structural coverage and fault spectrum overlap are comprehensively considered to select the final point as the optimal point; thereby improving the monitoring effect and achieving high-precision monitoring.

[0008] Optionally, determining the location of the sound source includes: determining a signal amplitude and a signal energy distribution according to the signal spectrum; Determining spectrum similarity based on the signal amplitude and a preset sound source spectrum library; determining an energy focus point according to the signal energy distribution and a preset pickup array position; Determining a separation distance according to the energy focus point and the signal attenuation; The sound source position is determined by combining the frequency spectrum similarity and the separation distance.

[0009] By adopting the above technical solution, the spectrum library is compared based on the signal amplitude and energy distribution to obtain spectrum similarity, and the sound source position is determined by combining the pickup array coordinates and attenuation. This shortens the positioning path and reduces environmental interference, ensuring rapid convergence of the sound source coordinates, maintaining high robustness and real-time performance under complex working conditions, and reducing interference.

[0010] Optionally, determining the energy focus point includes: Calculating the energy value of the signal collected by each sound pickup unit based on the signal energy distribution; According to the preset pickup array position coordinates, a mapping coordinate system of energy and position is established; Performing spatial focusing calculation on the energy value based on the mapping coordinate system to determine a spatial spectrum of energy distribution; Identifying an energy peak point based on the spatial spectrum, and calculating a sound pickup distance based on the energy peak point and a preset array of sound pickup units; The energy focus point is determined by combining the preset signal attenuation parameter and the sound pickup distance.

[0011] By adopting the above technical solution, an energy and position mapping coordinate system is constructed, and a spatial spectrum is obtained through spatial focusing calculation. After picking the peak point, the energy focusing point is locked in combination with the attenuation parameter. By replacing the traditional traversal with the spatial spectrum, the amount of calculation is significantly compressed, taking into account both positioning accuracy and computational efficiency.

[0012] Optionally, after matching and determining the noise reduction level at each position, the method further includes: determining waveform characteristics and spectrum envelope based on the signal spectrum; Determining a waveform matching degree based on the waveform characteristics and a preset waveform template library; determining a voiceprint parameter set based on the spectrum envelope; Determining a voiceprint feature by optimizing the voiceprint parameter set based on the waveform feature; Calculate the similarity between the voiceprint feature and the preset device standard voiceprint library to determine the voiceprint matching degree; Optimizing the voiceprint feature based on the voiceprint feature matching degree to update the voiceprint feature; The collected sound signal is subjected to noise reduction processing based on the updated voiceprint feature and the noise reduction level.

[0013] By adopting the above technical solution, waveform features and spectral envelopes are extracted, a voiceprint parameter set is generated, and iterative optimization is performed using a standard voiceprint library, followed by noise reduction. This converts acoustic features into quantifiable voiceprint indicators, achieves adaptive noise suppression, and provides continuous protection for high-fidelity collection of monitoring data.

[0014] Optionally, determining the voiceprint parameter set includes: performing preprocessing based on the sound signal to determine a noise reduction signal; Performing time-frequency analysis on the noise reduction signal to extract time domain features, frequency domain features and cepstrum features; Determining voiceprint feature parameters based on the extracted time domain features, the frequency domain features, and the cepstrum features; Comparing the spectrum envelope with preset voiceprint feature parameters to calculate envelope similarity; An operating weight parameter is introduced to perform weighted fusion on the waveform matching degree and the envelope similarity to determine a voiceprint parameter set.

[0015] By adopting the above technical solution, integrating time domain, frequency domain, and cepstrum features, introducing operating weights to weight waveform matching and envelope similarity, a voiceprint parameter set is constructed; by improving the completeness of feature expression, the sensitivity of voiceprints to minor faults is enhanced, thereby maintaining high recognition in multiple working condition noise environments.

[0016] Optionally, determining the waveform characteristics and the spectrum envelope includes: determining a signal amplitude and a signal period based on the signal spectrum; Calculating a phase difference according to the signal amplitude; Determining a frequency period segment according to the signal period and a preset period error range; Determining waveform characteristics in combination with the phase difference and the frequency period; Determining a time domain matrix based on the waveform characteristics, and determining an amplitude sequence according to the time domain matrix; After the amplitude sequence is smoothed, the spectrum envelope is determined in combination with the signal period.

[0017] By adopting the above technical solution, the phase difference and frequency period segment are extracted according to the signal amplitude and period, and the time domain matrix and amplitude sequence are generated, which are then smoothed to form the spectrum envelope; thus, the signal morphology is characterized in a matrix manner, ensuring high consistency between the waveform and the envelope, and laying a solid foundation for subsequent feature comparison.

[0018] Optionally, noise reduction processing includes: Determining a characteristic parameter sequence based on the voiceprint feature, and obtaining a preliminary curve by curve fitting; determining a trend change rate of the signal based on the preliminary curve; When the trend change rate exceeds a preset threshold, a fitting curve is determined in combination with a preset fitting correction coefficient; Calculating the energy proportion of noise based on the fitting curve; determining a noise interference level according to the energy proportion and the sound signal; When the noise interference level is greater than the noise reduction level, determining an over-subtraction parameter based on the noise interference level, and determining a smoothing parameter based on the noise reduction level; Noise reduction is performed according to the over-subtraction parameter and the smoothing parameter.

[0019] By adopting the above technical solution, the voiceprint characteristic curve is fitted, the trend change rate and the noise energy ratio are calculated, and the over-subtraction parameter and smoothing parameter are adjusted accordingly to complete the noise reduction; thereby achieving dynamic threshold control, avoiding signal distortion caused by excessive noise reduction, and retaining key fault information in a strong noise environment.

[0020] Optionally, noise reduction may include: Determining an effective signal bandwidth based on the signal spectrum and a preset bandwidth width; Converting the sound signal into the frequency domain to obtain a frequency band distribution; Identifying frequencies outside a preset frequency range based on the effective signal bandwidth, marking them and defining them as noise distribution; determining a noise ratio according to the frequency band distribution and the noise distribution; When the noise ratio is greater than a preset reference ratio, adjusting the voiceprint parameter set; A fitting model is generated based on the voiceprint parameter set and the waveform matching degree, and an effective signal after noise separation is reconstructed to complete noise reduction.

[0021] By adopting the above technical solution, the effective signal bandwidth is delineated, the frequency range exceeding the range is marked as noise distribution, and the effective signal after separation is reconstructed; the noise is accurately removed in the frequency domain to maintain signal integrity, thereby significantly improving the signal-to-noise ratio and monitoring accuracy in high-frequency interference scenarios.

[0022] Optionally, generating a fitted model includes: Constructing an initial feature set based on the voiceprint parameter set and the waveform matching degree; Performing time series segmentation on the initial feature set and calculating statistical features; Performing polynomial fitting on the statistical characteristics using the least squares method to determine a preliminary fitting curve; During the fitting process, when the noise interference level is less than the preset interference level, a preliminary fitting curve is obtained by direct fitting; When the noise interference level is not less than a preset interference level, filtering the statistical features is first performed, and then fitting is performed based on the filtered statistical features to obtain a preliminary fitting curve; When the trend change rate is greater than a preset ratio, adjusting the trend change rate to a modified trend change rate; A fitting model is generated based on the modified trend change rate and the preliminary fitting curve.

[0023] By adopting the above technical solution, the initial feature set is constructed and the time series is segmented. The least squares polynomial fitting is used, supplemented by filtering or trend correction to generate the final fitting model. By coupling statistical features with trends for modeling, the model generalization ability is improved, providing stable algorithm support for long-term tracking of equipment health status.

[0024] In a second aspect, the present application provides a system for selecting optimal monitoring points for equipment health, which adopts the following technical solutions: A system for selecting optimal monitoring points for equipment health, comprising: An acquisition module, used for acquiring sound signals; A memory for storing a program for selecting an optimal monitoring point for any equipment health; The processor loads and executes the program in the memory.

[0025] In summary, this application includes at least one of the following beneficial technical effects: 1. By adopting the above technical solution, sound signals are collected and the signal spectrum and attenuation are extracted. The signal-to-noise ratio and noise reduction level are calculated based on the spectrum. The final point is selected as the optimal point by combining the sound source location, operating condition fluctuation rate, structural coverage, and fault spectrum overlap. This allows the monitoring point to be accurately converged from a large number of candidate options, significantly improving the pertinence and reliability of subsequent fault diagnosis, thereby improving monitoring effectiveness and achieving high-precision monitoring. 2. By adopting the above technical solution, waveform features and spectral envelopes are extracted to generate a voiceprint parameter set, which is then iteratively optimized using a standard voiceprint library, followed by noise reduction. By converting acoustic features into quantifiable voiceprint indicators, adaptive noise suppression is achieved, providing continuous assurance for high-fidelity acquisition of monitoring data. 3. By adopting the above technical solution, an initial feature set is constructed and the time series is segmented. Least squares polynomial fitting is used, supplemented by filtering or trend correction to generate the final fitting model. By coupling statistical features with trend modeling, the model's generalization ability is improved, providing stable algorithm support for long-term tracking of equipment health status. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of a method for selecting optimal monitoring points for equipment health according to an embodiment of the present invention; Figure 2 is a flow chart of a noise reduction method according to an embodiment of the present invention; Figure 3 4 is a flow chart of a method for generating a fitting model according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to 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.

[0028] The embodiment of the present application discloses a method for selecting optimal monitoring points for equipment health.

[0029] Reference Figure 1 A method for selecting optimal monitoring points for equipment health includes the following steps: Step S100: collecting sound signals from preset monitoring points around the monitored device.

[0030] The monitored equipment refers to the machines that need to be monitored by sound pickup to identify the health of the equipment, such as motors, water pumps, etc., which are pre-set monitoring objects.

[0031] Monitoring points refer to the locations of pre-arranged pickup sensors around the monitored equipment.

[0032] The sound pickup sensor is a sensor that can collect the sound of the monitored equipment. It has array-arranged sound pickup units. The sound pickup unit refers to a single acoustic sensor. The monitoring point position and the sound pickup sensor are pre-set by technicians according to actual conditions and will not be described in detail here.

[0033] Sound signals refer to acoustic data from multiple monitoring points obtained by sound pickup sensors.

[0034] Step S101: determining a signal spectrum and a signal attenuation based on the sound signal.

[0035] Signal spectrum refers to the spectrum graph data of a sound signal in the frequency domain.

[0036] Signal attenuation refers to the energy loss of sound during propagation.

[0037] After the sound signal is converted into a digital signal through an analog-to-digital converter, the digital signal is converted into a sound spectrum diagram using fast Fourier transform, which is the signal spectrum.

[0038] The locations of various monitoring points are different, and the sound signals received are different. There is a distance difference between the monitoring points, and the sound signal also attenuates as the distance decreases. The sound signal and the monitoring point position of each monitoring point are input into the preset attenuation database, and the signal attenuation is obtained by matching. The attenuation database is a database pre-set by technical personnel based on actual conditions. The attenuation database is provided with a formula for determining the signal attenuation based on the monitoring point position and the corresponding sound signal. The actual formula is pre-set by technical personnel based on actual conditions and will not be elaborated here.

[0039] Step S102: determining the sound source position according to the signal spectrum and the signal attenuation.

[0040] The sound source location refers to the coordinates of the sound source generated by equipment noise or abnormal vibration.

[0041] During sound propagation, it attenuates due to factors like diffusion and absorption, and the amount of attenuation is directly related to the propagation distance. The distance can be inferred from the attenuation of the sound source; the greater the distance, the more dispersed the sound energy distribution. Substituting the signal spectrum and signal attenuation into a preset sound source formula yields the corresponding sound source location. The sound source formula is pre-set and will not be detailed here.

[0042] Step S103: Based on the signal spectrum, the signal-to-noise ratio of the preset monitoring point signal is calculated, the noise reduction level of each position is matched and determined, and a comprehensive score is obtained by weighting the noise reduction level and the sound source position.

[0043] The signal-to-noise ratio refers to the energy ratio of the sound signal to the noise.

[0044] The noise reduction level refers to the noise reduction intensity classification required for the corresponding position. The noise reduction intensity corresponding to each level is pre-set by technical personnel based on actual conditions and will not be elaborated here.

[0045] The comprehensive score refers to the score used to judge the monitoring effect of the monitoring point location. It is obtained by multiplying the sound source location and noise reduction level by the preset score weight. The score weight is pre-set by technical personnel based on actual conditions and will not be elaborated here.

[0046] The preset monitoring point signal is identified in the signal spectrum, and the ratio between the monitoring point signal and the noise other than the monitoring point signal in the signal spectrum is calculated as the signal-to-noise ratio.

[0047] Different signal-to-noise ratios correspond to different noise reduction levels. A correspondence between the signal-to-noise ratio and the noise reduction level is pre-stored. The corresponding noise reduction level can be determined based on the obtained signal-to-noise ratio. The correspondence between the signal-to-noise ratio and the noise reduction level is pre-set by technicians based on actual conditions and will not be elaborated here.

[0048] Step S104: Calculate the fluctuation rate under various working conditions based on the monitoring points and the corresponding sound signals, and select locations where the fluctuation rate is less than a preset stability threshold and the comprehensive score is higher than a preset score threshold as secondary monitoring points.

[0049] Fluctuation rate refers to the rate of change of the sound signal amplitude under different working conditions and is used to judge the stability of the sound signal.

[0050] The preset stability threshold refers to the maximum allowable fluctuation range, which is pre-set by technical personnel based on actual conditions and is not described in detail here.

[0051] The scoring threshold refers to the score line used to determine whether the comprehensive score of the monitoring point is qualified. It is pre-set by technical personnel based on actual conditions and will not be elaborated here.

[0052] Different sound data under various working conditions are pre-set, and the sound signals at different monitoring points are compared with the corresponding sound data to calculate the difference between the corresponding sound signals and sound data under the same working conditions, and then determine the corresponding fluctuation rate with the sound data.

[0053] For example, the parameter of the sound data is 10, and the parameter obtained by the sound signal is 9. The difference between the two is 10-9=1, so the corresponding volatility is 1 / 10*100%=10%.

[0054] Step S105: determining the location distribution of the secondary monitoring points according to the sound source location and the preset equipment structure.

[0055] The equipment structure refers to the positional composition of the various parts of the monitored equipment, which is pre-set by technical personnel based on actual conditions and will not be described in detail here.

[0056] When the monitoring point is set, the position is known. The position of the secondary monitoring point on the equipment structure is determined in combination with the equipment structure. The distribution of the secondary monitoring points around the sound source position is determined according to the sound source position.

[0057] For example, suppose the coordinates of the sound source position determined in the initial state (first coordinate system) are (1, 2), and the coordinates of one of the secondary monitoring points are (3, 5), so that the coordinates of the sound source position are used as the origin of the coordinate system, and the coordinates of the secondary monitoring point (2, 3) are determined with the sound source position as the reference point (second coordinate system). The equipment structure is known, the distance between the positions is known, and the coverage of the coverage equipment at each position can also be determined, thereby obtaining the distribution of the secondary monitoring points at the sound source position.

[0058] Step S106: selecting a point with the largest coverage from the position distribution, and determining a point with the highest overlap as the optimal point based on matching the signal spectrum with a preset fault spectrum.

[0059] The fault spectrum refers to the spectrum characteristics of a device failure. It is pre-set by technicians based on actual conditions and is not described here in detail.

[0060] Coverage refers to the richness of the types of sound signals that can be collected by the monitoring point at that location. The coverage of each location can be determined based on the analysis of the sound signal on the signal spectrum.

[0061] Coincidence refers to the similarity between the signal spectrum and the preset fault spectrum, and is used to screen for the optimal point. The signal spectrum and the fault spectrum are compared and analyzed for matching in terms of frequency content, amplitude distribution, harmonic characteristics, and other dimensions. The optimal point is then selected, representing the point in the sound signal with the most complete fault spectrum.

[0062] Determining the location of a sound source involves the following steps: Step S200: determining the signal amplitude and signal energy distribution according to the signal spectrum.

[0063] Signal amplitude refers to the maximum fluctuation range in the spectrum over a period of time.

[0064] Signal energy distribution refers to the distribution of energy in the frequency domain.

[0065] From the spectrum graph data of the signal spectrum, identify the frequency point with the largest amplitude at each frequency in the spectrum, record its amplitude, and thus obtain the signal amplitude.

[0066] The signal energy distribution E(f) is proportional to the square of the signal amplitude A(f). If the spectrum is the amplitude spectrum A(f), the energy spectrum can be obtained by square operation: (k is a preset constant related to the signal sampling parameters).

[0067] Step S201: determining spectrum similarity according to the signal amplitude and a preset sound source spectrum library.

[0068] The sound source spectrum library refers to a database of pre-stored characteristic amplitudes of various sound sources, which is pre-set by technical personnel based on actual conditions and will not be described in detail here.

[0069] Spectral similarity refers to the degree of match between the signal amplitude and the characteristic amplitude of the sound source in the sound source spectrum library, and is used to judge changes in the propagation of sound signals.

[0070] The amplitude of each frequency point on the signal amplitude is compared with the characteristic amplitudes of various sound sources pre-stored in the sound source spectrum library to determine the degree of coincidence between the amplitudes, and the spectrum similarity is obtained after weighting.

[0071] Step S202: determining an energy focus point according to the signal energy distribution and a preset pickup array position.

[0072] The sound pickup array position refers to the spatial coordinates of the arrangement of the sound pickup sensor array, which is pre-set by technical personnel based on actual conditions and will not be described in detail here.

[0073] The energy focusing point refers to the position where energy converges and overlaps the most in space.

[0074] As sound propagates through space, its energy attenuates with distance. By analyzing the differences in signal energy received by each pickup unit in the pickup array, the energy received by each pickup unit is correlated with its distance from the energy focal point. Combined with the known positions of the pickup units, the spatial coordinates where the sound energy is most concentrated are determined, representing the energy focal point. The spatial coordinates of the energy concentration are thus obtained. This method is not described in detail here. See steps S300 to S304 for the specific method.

[0075] Step S203: determining the separation distance according to the energy focus point and the signal attenuation.

[0076] Standoff distance refers to the physical distance from the energy focus point to the pickup sensor.

[0077] Since the energy of sound signals decays with distance during propagation, the distance calculated by inputting the signal attenuation into the sound source formula based on the energy focal point is the separation distance.

[0078] Step S204: Determine the sound source position by combining the frequency spectrum similarity and the interval distance.

[0079] The spectrum similarity is used to exclude those with low similarity, and only the focal points that match the sound sources with high similarity are retained. At the same time, the position of each unit of the pickup array is known. Combined with the calculated interval distance, the coordinate range of the focal point can be narrowed through spatial geometric positioning, so that the sound source position can be determined according to the interval distance, thereby reducing interference.

[0080] Determining the energy focus point involves the following steps: Step S300: Calculating the energy value of the signal collected by each sound pickup unit based on the signal energy distribution.

[0081] The approximate energy situation of each pickup unit is preliminarily determined based on the signal energy distribution, and then the final energy value is obtained by weighted calculation with the signal energy value collected by each unit. The weighted calculation method is common knowledge among those skilled in the art and will not be described in detail here.

[0082] Step S301: establishing a mapping coordinate system of energy and position according to the preset position coordinates of the sound pickup array.

[0083] The position coordinates of the sound pickup array refer to the coordinate data of each sound pickup unit array, which are pre-set by technical personnel based on actual conditions and will not be described in detail here.

[0084] Mapping coordinates refer to the mathematical model that maps energy values ​​to spatial coordinates.

[0085] According to the energy value of each pickup unit and the position of the pickup array, the energy and the corresponding position are matched one by one to establish a mapping coordinate system of energy and position. The specific method is common knowledge to those skilled in the art and will not be described here.

[0086] Step S302: performing spatial focusing calculation on the energy value based on the mapping coordinate system to determine a spatial spectrum of energy distribution.

[0087] Spatial spectrum refers to the image of energy distribution in space.

[0088] According to the energy value of the pickup array, mathematical calculations are performed to infer the corresponding points in space where the energy is emitted by different pickup units. Each point is marked in space. Different energy values ​​indicate different points. The positions of the points in space are converted into intuitive images to obtain a spatial spectrum.

[0089] Step S303: obtaining energy peak points based on the spatial spectrum recognition, and calculating sound pickup distances based on the energy peak points and preset sound pickup unit arrays.

[0090] Each sound pickup unit array refers to the arrangement and distance between each sound pickup unit array, which is pre-set by technical personnel according to actual conditions and will not be described in detail here.

[0091] The energy peak point refers to the position in the spectrum where the overlap of focused energy points is the highest.

[0092] The pickup distance refers to the distance from the peak point to the sensor.

[0093] Based on the spatial spectrum, the peak points are extracted by MATLAB software to obtain the energy peak points. The specific method is common knowledge to those skilled in the art and will not be described in detail here.

[0094] The pickup distance is calculated by calculating the distance between the point of the pickup unit array and the energy peak point. For example, the pickup unit array point is (X2, Y2, Z2), and the energy peak point is (X1, Y1, Z1). The calculated pickup distance L = √((X1-X2) 2 + (Y1-Y2) 2 + (Z1-Z2) 2 ).

[0095] Step S304: Determine the energy focus point by combining the preset signal attenuation parameter and the sound pickup distance.

[0096] The signal attenuation parameter refers to the mathematical relationship between signal attenuation and distance, which is preset by technicians based on actual conditions and will not be described in detail here.

[0097] The energy peak point may include false points caused by environmental interference and needs to be verified by the signal attenuation parameter. Therefore, the signal attenuation parameter and the pickup distance are used to determine whether the energy peak point meets the signal attenuation parameter. If it meets the signal attenuation parameter, the energy peak point is the energy focus point. If it does not meet the signal attenuation parameter, the energy peak point is corrected according to the signal attenuation parameter and the pickup distance to obtain the energy focus point.

[0098] After matching and determining the noise reduction level at each location, the following steps are included: Step S400: determining waveform characteristics and spectrum envelope based on the signal spectrum.

[0099] Waveform characteristics refer to the shape and periodic characteristics of the sound signal's wave fluctuations in the time domain.

[0100] The spectrum envelope refers to the envelope of the signal spectrum, which is used to describe the distribution trend of energy with frequency.

[0101] The specific method of determining the waveform characteristics and the spectrum envelope based on the signal spectrum refers to steps S600 to S605 and is not described in detail here.

[0102] Step S401: determining a waveform matching degree according to the waveform characteristics and a preset waveform template library.

[0103] The waveform template library refers to a database that stores standard waveform features, which is pre-set by technical personnel based on actual conditions and will not be described in detail here.

[0104] Waveform matching refers to the similarity between the current waveform characteristics and the standard waveform in the template library.

[0105] The waveform features are compared with the same standard waveform features in the waveform template library to obtain the waveform matching degree.

[0106] Step S402: Determine a voiceprint parameter set based on the spectrum envelope.

[0107] The voiceprint parameter set refers to a set of parameter vectors extracted from the spectrum envelope for voiceprint recognition.

[0108] The specific method for determining the voiceprint parameter set is referred to steps S500 to S504 and will not be described in detail here.

[0109] Step S403: Based on the waveform characteristics, determine the voiceprint characteristics by optimizing the voiceprint parameter set.

[0110] Voiceprint features refer to the voiceprint data of the target signal, which is used to match fault sounds and identify core fault conditions of the device.

[0111] The time-domain morphological features of the waveform (such as the waveform details corresponding to the period, peak value, and spectrum envelope) are modified through the parameters in the voiceprint parameter set to form a voiceprint feature that is both stable and distinctive.

[0112] Step S404: Calculate the similarity between the voiceprint feature and a preset device standard voiceprint library to determine the voiceprint matching degree.

[0113] The device standard voiceprint library refers to a database that stores the voiceprint characteristics of health devices. It is pre-set by technical personnel based on actual conditions and will not be described in detail here.

[0114] Voiceprint matching refers to the similarity between the current voiceprint feature and the standard voiceprint library.

[0115] The voiceprint feature is similarly calculated with the standard voiceprint matched in the device standard voiceprint library to determine the voiceprint matching degree.

[0116] Step S405: Optimizing the voiceprint feature based on the voiceprint feature matching degree to update the voiceprint feature.

[0117] By analyzing the degree of matching between the current voiceprint features and the standard voiceprint, the deviation components in the positioning features that lead to low matching degree are identified. Based on the screened deviation parameters, the corresponding features of the current voiceprint are directly corrected to obtain a more accurate voiceprint feature.

[0118] Step S406: performing noise reduction processing on the collected sound signal based on the updated voiceprint feature and the noise reduction level.

[0119] The specific method of the noise reduction processing is referred to steps S700 to S805 and will not be described in detail here.

[0120] Determining the voiceprint parameter set includes the following steps: Step S500: Preprocessing is performed based on the sound signal to determine a noise reduction signal.

[0121] Preprocessing refers to operations such as filtering and normalization on sound signals. Preprocessing methods are common knowledge among those skilled in the art and will not be described in detail here.

[0122] A noise reduction signal is obtained through preprocessing to reduce noise interference.

[0123] Step S501: performing time-frequency analysis on the noise reduction signal to extract time domain features, frequency domain features and cepstrum features.

[0124] Time domain features refer to the characteristics of the signal in the time dimension, such as amplitude, period, etc.

[0125] Frequency domain features refer to the characteristics of the signal in the frequency dimension, such as spectrum energy distribution.

[0126] The cepstrum feature refers to the inverse transform feature of the logarithmic spectrum of a signal.

[0127] By analyzing the dynamic changes of the signal in the time domain and frequency domain, the non-stationary characteristics of the signal are fully captured, and the time domain, frequency domain and cepstrum features are extracted from it. The extraction method is common knowledge among those skilled in the art and will not be described here.

[0128] Step S502: Determine voiceprint feature parameters based on the extracted time domain features, the frequency domain features, and the cepstrum features.

[0129] The voiceprint feature parameters refer to a set of parameters obtained from the time domain features, the frequency domain features and the cepstrum features.

[0130] Time domain features reflect the dynamic amplitude changes of the sound. They provide statistical characteristics of short-term energy and peak intervals, as well as the mean zero-crossing rate. Frequency domain features reflect the frequency composition of the sound. They provide statistical characteristics of the fundamental frequency, formant characteristics, and the mean of the spectral entropy. Cepstrum features effectively capture the macroscopic characteristics of vocal tract resonance. They provide the static and dynamic MFCC coefficients and the cepstrum mean.

[0131] A variety of features are obtained from the time domain features, the frequency domain features and the cepstrum features. Voiceprint feature parameters are formed based on the obtained multiple features. The specific method is common knowledge to those skilled in the art and will not be described here.

[0132] Step S503: Compare the spectrum envelope with preset voiceprint feature parameters to calculate envelope similarity.

[0133] Envelope similarity refers to the similarity between the spectrum envelope and the standard voiceprint feature parameters.

[0134] After extracting the corresponding envelope from the voiceprint feature parameters, the similarity between the spectrum envelope and the extracted envelope is calculated to obtain the envelope similarity.

[0135] Step S504: introducing an operating weight parameter, performing weighted fusion on the waveform matching degree and the envelope similarity, and determining a voiceprint parameter set.

[0136] The operating weight parameter refers to the weight ratio used to adjust the waveform matching and envelope similarity, which is pre-set by the technical staff and will not be described in detail here.

[0137] The waveform matching degree and envelope similarity are weighted averaged according to the running weight parameter to obtain the voiceprint parameter set.

[0138] Determining the waveform characteristics and spectral envelope involves the following steps: Step S600: determining a signal amplitude and a signal period based on the signal spectrum.

[0139] Signal amplitude refers to the peak magnitude of the signal waveform.

[0140] The signal period refers to the time interval during which the signal waveform repeats.

[0141] In the spectrum diagram data of the signal spectrum, the amplitude of the peak value on the extracted signal waveform is the signal amplitude; the signal period is obtained by the time interval of the repeated signal waveform. The determination method is common knowledge to those skilled in the art and will not be elaborated here.

[0142] Step S601: Calculate the phase difference according to the signal amplitude.

[0143] Phase difference refers to the phase offset between signal waveforms.

[0144] The phase difference is derived based on the peak amplitude of the signal amplitude and the mathematical relationship between the signal amplitude and the phase difference. The calculation method is common knowledge to those skilled in the art and will not be described in detail here.

[0145] Step S602: determining a frequency period segment according to the signal period and a preset period error range.

[0146] The cycle error range refers to the allowable cycle deviation range, which is used to define the cycle range to avoid the cycle being too large. It is pre-set by technical personnel and will not be described in detail here.

[0147] The frequency cycle segment refers to the valid range of the signal frequency.

[0148] During the signal cycle, signals outside the periodic error range are deleted, and the corresponding continuous frequency interval signals that meet the periodic stability requirements are divided in the frequency domain. The periodic error range is used to filter stable periodic signal components and eliminate unstable frequency components caused by excessive periodic fluctuations.

[0149] Step S603: Determine waveform characteristics by combining the phase difference and the frequency period segment.

[0150] The frequency period segment determines the stable frequency components of the signal, and then the phase difference between these components is used to further obtain the morphological properties of the waveform, which is the waveform feature.

[0151] Step S604: determining a time domain matrix based on the waveform characteristics, and determining an amplitude sequence according to the time domain matrix.

[0152] The time domain matrix is ​​a matrix form in which signals are stored in a structured manner in the time dimension.

[0153] Amplitude sequence refers to the sequence of signal amplitude changes over time.

[0154] The time domain features of the waveform are converted into matrix data through time window segmentation and feature extraction, and then the amplitude information that changes with time is extracted from the matrix to form an ordered amplitude sequence. The specific method is common knowledge among those skilled in the art and will not be described here.

[0155] Step S605: After smoothing the amplitude sequence, the spectrum envelope is determined in combination with the signal period.

[0156] Smoothing refers to filtering the amplitude sequence. The smoothing method is common knowledge to those skilled in the art and will not be described in detail here.

[0157] The spectrum envelope is the overall profile of the amplitude variation with frequency in the frequency domain. Based on the smoothed amplitude sequence and signal period, the time domain is first converted to the frequency domain through Fourier transform, and then the converted signal frequency domain is matched with periodic features to extract the spectrum envelope. The determination method is common knowledge among technicians in this field and will not be elaborated here.

[0158] The noise reduction process includes the following steps: Step S700: determining a characteristic parameter sequence based on the voiceprint feature, and obtaining a preliminary curve through curve fitting.

[0159] The feature parameter sequence refers to the parameter sequence extracted from the voiceprint features and is used to fit the noise reduction curve.

[0160] The preliminary curve refers to the initial noise reduction curve obtained by fitting, which is convenient for subsequent correction.

[0161] Relevant parameters such as basic frequency and cepstrum features are extracted from the voiceprint features, and then the discrete relevant parameters are input into the fitting software through mathematical fitting methods, thereby converting them into a continuous curve to obtain a preliminary curve.

[0162] Step S701: Determine the trend change rate of the signal according to the preliminary curve.

[0163] The trend rate of change refers to the rate at which a signal changes over time.

[0164] Divide the data of each point of the preliminary curve by the corresponding time to obtain the rate of change of the trend, which is the trend change rate.

[0165] Step S702: When the trend change rate exceeds a preset threshold, a fitting curve is determined in combination with a preset fitting correction coefficient.

[0166] The preset threshold refers to a critical value at which the trend change rate needs to be corrected, which is pre-set by technical personnel and will not be described in detail here.

[0167] The fitting correction coefficient refers to the coefficient used to correct the fitting curve.

[0168] Based on the trend change rate, the interval in the positioning curve where the change rate exceeds the preset threshold is located, and the fitting point data in this interval is multiplied by the fitting correction parameter to ensure the overall continuity of the curve and the fit with the data.

[0169] When the trend change rate exceeds the preset threshold, it indicates that the trend change of the current fitting curve is too steep, or the curve is distorted due to noise interference or abnormal data fluctuations. The out-of-range values ​​in the preliminary curve are corrected to ensure the reliability of the fitting curve.

[0170] Step S703: Calculate the energy ratio of noise based on the fitting curve.

[0171] Energy ratio refers to the ratio of noise energy to total signal energy.

[0172] The fitting residual is further processed by fitting the curve in combination with filtering or transformation algorithms (such as wavelet transform, FFT) to separate the target signal and the noise signal, and then the energy of the two is calculated separately. Finally, the proportion of noise energy in the total signal energy is obtained, which is the energy proportion. The specific method is common knowledge among technicians in this field and will not be described here.

[0173] Step S704: Determine a noise interference level according to the energy ratio and the sound signal.

[0174] Noise interference level refers to the severity of noise interference to the signal.

[0175] Based on the sound signal, the greater the energy ratio, the corresponding noise interference level is greater. The energy ratio and the sound signal are input into the preset noise interference level database, and the noise interference level is obtained by matching. The noise interference level database is a database pre-set by technical personnel according to actual conditions. The noise interference level database is provided with a comparison table for determining the noise interference level according to the energy ratio and the sound signal matching. The actual comparison table is pre-set by technical personnel according to actual conditions and will not be elaborated here.

[0176] Step S705: When the noise interference level is greater than the noise reduction level, determining an over-subtraction parameter based on the noise interference level, and determining a smoothing parameter according to the noise reduction level.

[0177] The over-reduction parameter refers to a parameter used for excessive noise reduction.

[0178] Smoothing parameters refer to the parameters used for smoothing noise reduction.

[0179] The noise interference level and the over-reduction parameter, the noise reduction level and the smoothing parameter are all positively correlated. The noise interference level and the noise reduction level are input into the corresponding mapping relationship database to match the corresponding over-reduction parameter and the smoothing parameter. The mapping relationship database pre-stores the mapping relationship between the noise interference level and the over-reduction parameter, and the mapping relationship between the noise reduction level and the smoothing parameter. The mapping relationship is pre-set by technical personnel according to actual conditions and will not be elaborated here.

[0180] Step S706: performing noise reduction according to the over-subtraction parameter and the smoothing parameter.

[0181] The obtained over-subtraction parameter and smoothing parameter are used to reduce noise on the sound signal by using spectral subtraction. Spectral subtraction is common knowledge to those skilled in the art and will not be described in detail here.

[0182] Noise reduction involves the following steps: Step S800: determining an effective signal bandwidth according to the signal spectrum and a preset bandwidth width.

[0183] The bandwidth refers to the width range of the set signal frequency band, which is used to extract and determine the effective signal bandwidth. It is pre-set by technical personnel and will not be described in detail here.

[0184] The effective signal bandwidth refers to the frequency band range that contains the target signal, which is used to define the frequency interval that contains the main useful signal energy.

[0185] In the signal spectrum, the signals that exceed the range are deleted according to the bandwidth width to obtain the effective signal bandwidth.

[0186] Step S801: convert the sound signal into the frequency domain to obtain frequency band distribution.

[0187] Frequency band distribution refers to the distribution of signal energy at signal frequencies in the frequency domain.

[0188] The sound signal is converted into the frequency domain through Fourier transform to obtain the frequency band distribution.

[0189] Step S802: Identify frequencies exceeding a preset frequency range based on the effective signal bandwidth, mark them, and define them as noise distribution.

[0190] The frequency range refers to the frequency boundary of the effective signal, which is pre-set by the technicians and will not be described in detail here.

[0191] Noise distribution refers to the frequency region marked as noise.

[0192] By using the frequency range to identify all frequency components whose frequency values ​​exceed the frequency range within the effective signal bandwidth, the excess frequencies are marked to determine the noise distribution.

[0193] Step S803: determining a noise ratio according to the frequency band distribution and the noise distribution.

[0194] The noise ratio refers to the ratio of the noise frequency to the signal frequency.

[0195] The noise ratio is calculated by dividing the noise frequency of the noise distribution by the signal frequency of the frequency distribution.

[0196] Step S804: When the noise ratio is greater than a preset reference ratio, adjust the voiceprint parameter set.

[0197] The reference ratio refers to the upper limit of the allowable noise ratio, which is a reference value used to judge whether the noise is too loud. It is pre-set by technical personnel and will not be described in detail here.

[0198] When the noise ratio is greater than the preset reference ratio, it means that the noise will interfere with the fault diagnosis and analysis. Step S805: generating a fitting model based on the voiceprint parameter set and the waveform matching degree, reconstructing a valid signal after noise separation, and completing noise reduction.

[0199] A fitting model is generated based on the voiceprint parameter set and the waveform matching degree. For a specific fitting method, refer to steps S900 to S906.

[0200] The noise is separated according to the fitting model to obtain the effective signal and complete the noise reduction.

[0201] Generating a fitted model involves the following steps: Step S900: constructing an initial feature set based on the voiceprint parameter set and the waveform matching degree.

[0202] The initial feature set value refers to the set of preliminary parameter features used to fit the generative model.

[0203] Multiple parameter features and waveform matching in the voiceprint parameter set are used as preliminary parameters to construct an initial feature set.

[0204] Step S901: performing time series segmentation on the initial feature set and calculating statistical features.

[0205] Statistical features refer to the parameter set obtained by statistically analyzing the signal over time.

[0206] The initial feature set is divided according to time order, so as to obtain the statistical features of each time period in the feature set in chronological order, and the original continuous time series data is converted into a point set feature vector, which is the statistical feature.

[0207] Step S902: performing polynomial fitting on the statistical features using the least squares method to determine a preliminary fitting curve.

[0208] A set of coefficients is found through the least squares method to minimize the sum of square errors between the predicted value of the fitting curve and the actual statistical characteristic value. The statistical characteristics are input into MATLAB software, and the characteristic lines are extracted based on the block characteristics of the point cloud. The parametric curve modeling (fitting regular geometric shapes) method is selected to generate the corresponding fitting curve.

[0209] Step S903: During the fitting process, when the noise interference level is less than the preset interference level, a preliminary fitting curve is directly obtained by fitting.

[0210] The preset interference level refers to the critical level of noise interference, which is used to determine whether noise interference will affect fitting. It is pre-set by technical personnel and will not be described in detail here.

[0211] When the noise interference level is less than the preset interference level, it means that the noise will not interfere with the fitting result, and the fitting is performed directly.

[0212] Step S904: When the noise interference level is not less than a preset interference level, the statistical features are first filtered, and then a preliminary fitting curve is obtained based on the filtered statistical features.

[0213] Filtering refers to filtering the signal to remove noise and reduce noise interference.

[0214] When the noise interference level is not less than the preset interference level, it means that the noise is too large, which will cause the fitting curve to be distorted. From the signal containing noise interference, the target frequency signal is retained and other unnecessary components (such as noise, clutter, interference frequency, etc.) are removed. After filtering, fitting is performed.

[0215] Step S905: When the trend change rate is greater than a preset ratio, the trend change rate is adjusted to a modified trend change rate.

[0216] The preset ratio refers to a critical ratio of the trend change rate, which is a benchmark value used to determine whether the trend change rate is too large. It is pre-set by technical personnel and will not be described in detail here.

[0217] The modified trend change rate refers to the adjusted trend change rate, which is obtained by multiplying the trend change rate by 1.2.

[0218] When the trend change rate is greater than the preset ratio, it indicates that the signal trend has a more obvious fluctuation or turning point. At this time, by expanding the trend change rate for targeted correction, the fitting model can follow the changes in the signal trend more quickly, reduce the fitting deviation caused by trend mutations, and enhance the model's ability to capture dynamic signals.

[0219] Step S906: generating a fitting model based on the modified trend change rate and the preliminary fitting curve.

[0220] The modified trend change rate is used as a weight factor and multiplied by the coefficient in the preliminary fitting curve to change the curvature of the preliminary fitting curve, thereby obtaining a fitting model so that it can be adjusted synchronously according to actual conditions and reduce deviations.

[0221] Based on the same inventive concept, an embodiment of the present invention provides a system for selecting optimal monitoring points for equipment health, including: The acquisition module is used to obtain the sound signal and the energy value of the signal.

[0222] A memory is used to store a program for selecting an optimal monitoring point for any equipment health.

[0223] The processor loads and executes the program in the memory.

[0224] Those skilled in the art will clearly understand that for the sake of convenience and brevity, the division of the above-mentioned functional modules is only used as an example for illustration. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working processes of the above-mentioned systems, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0225] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for selecting optimal monitoring points for equipment health, characterized in that: include: Collect sound signals from preset monitoring points around the monitored equipment; determining a signal spectrum and a signal attenuation amount based on the sound signal; determining a sound source position according to the signal spectrum and the signal attenuation; Based on the signal spectrum, the signal-to-noise ratio of the preset monitoring point signal is calculated, the noise reduction level of each position is matched and determined, and a comprehensive score is obtained according to the weighting of the noise reduction level and the sound source position; Calculating a comprehensive score of the monitoring point based on the sound source location and the noise reduction level; Calculating the fluctuation rates under various working conditions based on the monitoring points and the corresponding sound signals, and selecting locations where the fluctuation rates are less than a preset stability threshold and the comprehensive scores are higher than a preset score threshold as secondary monitoring points; Determining the position distribution of the secondary monitoring points according to the sound source position and the preset equipment structure; A point with the largest coverage is selected from the position distribution, and a point with the highest overlap is determined as the optimal point based on matching between the signal spectrum and a preset fault spectrum.

2. The method for selecting the optimal monitoring point for equipment health according to claim 1, characterized in that: Determining the location of a sound source involves: determining a signal amplitude and a signal energy distribution according to the signal spectrum; Determining spectrum similarity based on the signal amplitude and a preset sound source spectrum library; determining an energy focus point according to the signal energy distribution and a preset pickup array position; Determining a separation distance according to the energy focus point and the signal attenuation; The sound source position is determined by combining the frequency spectrum similarity and the separation distance.

3. The method for selecting the optimal monitoring point for equipment health according to claim 2, characterized in that: Determining the energy focus includes: Calculating the energy value of the signal collected by each sound pickup unit based on the signal energy distribution; According to the preset pickup array position coordinates, a mapping coordinate system of energy and position is established; Performing spatial focusing calculation on the energy value based on the mapping coordinate system to determine a spatial spectrum of energy distribution; Identifying an energy peak point based on the spatial spectrum, and calculating a sound pickup distance based on the energy peak point and a preset array of sound pickup units; The energy focus point is determined by combining the preset signal attenuation parameter and the sound pickup distance.

4. The method for selecting the optimal monitoring point for equipment health according to claim 1, characterized in that: After matching to determine the noise reduction level at each location, it also includes: determining waveform characteristics and spectrum envelope based on the signal spectrum; Determining a waveform matching degree based on the waveform characteristics and a preset waveform template library; determining a voiceprint parameter set based on the spectrum envelope; Determining a voiceprint feature by optimizing the voiceprint parameter set based on the waveform feature; Calculate the similarity between the voiceprint feature and the preset device standard voiceprint library to determine the voiceprint matching degree; Optimizing the voiceprint feature based on the voiceprint feature matching degree to update the voiceprint feature; The collected sound signal is subjected to noise reduction processing based on the updated voiceprint feature and the noise reduction level.

5. The method for selecting the optimal monitoring point for equipment health according to claim 4, characterized in that: Determine the voiceprint parameter set including: performing preprocessing based on the sound signal to determine a noise reduction signal; Performing time-frequency analysis on the noise reduction signal to extract time domain features, frequency domain features and cepstrum features; Determining voiceprint feature parameters based on the extracted time domain features, the frequency domain features, and the cepstrum features; Comparing the spectrum envelope with preset voiceprint feature parameters to calculate envelope similarity; An operating weight parameter is introduced to perform weighted fusion on the waveform matching degree and the envelope similarity to determine a voiceprint parameter set.

6. The method for selecting the optimal monitoring point for equipment health according to claim 4, characterized in that: Determining waveform characteristics and spectral envelope includes: determining a signal amplitude and a signal period based on the signal spectrum; Calculating a phase difference according to the signal amplitude; Determining a frequency period segment according to the signal period and a preset period error range; Determining waveform characteristics in combination with the phase difference and the frequency period; Determining a time domain matrix based on the waveform characteristics, and determining an amplitude sequence according to the time domain matrix; After the amplitude sequence is smoothed, the spectrum envelope is determined in combination with the signal period.

7. The method for selecting the optimal monitoring point for equipment health according to claim 4, characterized in that: Noise reduction processing includes: Determine a characteristic parameter sequence based on the voiceprint feature, and obtain a preliminary curve by curve fitting; determining a trend change rate of the signal based on the preliminary curve; When the trend change rate exceeds a preset threshold, a fitting curve is determined in combination with a preset fitting correction coefficient; Calculating the energy proportion of noise based on the fitting curve; determining a noise interference level according to the energy proportion and the sound signal; When the noise interference level is greater than the noise reduction level, determining an over-subtraction parameter based on the noise interference level, and determining a smoothing parameter based on the noise reduction level; Noise reduction is performed according to the over-subtraction parameter and the smoothing parameter.

8. The method for selecting the optimal monitoring point for equipment health according to claim 7, characterized in that: Noise reduction includes: Determining an effective signal bandwidth based on the signal spectrum and a preset bandwidth width; Converting the sound signal into the frequency domain to obtain a frequency band distribution; Identifying frequencies outside a preset frequency range based on the effective signal bandwidth, marking them and defining them as noise distribution; determining a noise ratio according to the frequency band distribution and the noise distribution; When the noise ratio is greater than a preset reference ratio, adjusting the voiceprint parameter set; A fitting model is generated based on the voiceprint parameter set and the waveform matching degree, and an effective signal after noise separation is reconstructed to complete noise reduction.

9. The method for selecting the optimal monitoring point for equipment health according to claim 8, characterized in that: Generating a fitted model involves: Constructing an initial feature set based on the voiceprint parameter set and the waveform matching degree; Performing time series segmentation on the initial feature set and calculating statistical features; Performing polynomial fitting on the statistical characteristics using the least squares method to determine a preliminary fitting curve; During the fitting process, when the noise interference level is less than the preset interference level, a preliminary fitting curve is obtained by direct fitting; When the noise interference level is not less than a preset interference level, filtering the statistical features is first performed, and then fitting is performed based on the filtered statistical features to obtain a preliminary fitting curve; When the trend change rate is greater than a preset ratio, adjusting the trend change rate to a modified trend change rate; A fitting model is generated based on the modified trend change rate and the preliminary fitting curve.

10. A system for selecting optimal monitoring points for equipment health, characterized in that: include: An acquisition module, used for acquiring sound signals; A memory for storing a program of a method for selecting an optimal monitoring point for equipment health according to any one of claims 1 to 9; The processor loads and executes the program in the memory.

Images