Fault voiceprint monitoring method and system for pump unit monitoring system

Abstract

The invention relates to the technical field of pump unit monitoring, and discloses a fault voiceprint monitoring method and system for a pump unit monitoring system, and the method comprises the steps: setting a sound collection device, and obtaining the environment sound in a monitoring region based on the sound collection device; acquiring a sound propagation path of each pump unit in the monitoring area, and determining a spatial position between the sound acquisition device and each pump unit according to the sound propagation path; according to the spatial position, obtaining the operation sound of each pump unit in the determined environment sound, obtaining the voiceprint feature in the operation sound, and determining whether the pump unit has an operation fault according to the relationship between the voiceprint feature and the preset voiceprint feature vector of the pump unit; and when it is determined that the pump unit has the operation fault, the fault type of the pump unit is determined according to the relationship between the voiceprint feature and each fault voiceprint feature. According to the invention, monitoring and diagnosis of the operation state and the fault type of the pump unit are realized through voiceprint recognition and positioning, and the efficiency and the accuracy of fault detection are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of pump unit monitoring, and in particular to a fault soundprint monitoring method and system for a pump unit monitoring system. Background Art

[0002] Pump units are important equipment in industrial production, urban water supply and drainage systems, and their operating status directly affects the stability and reliability of the entire system. Since pump units operate in complex environments for a long time, they are easily affected by factors such as vibration, pressure fluctuations, and fluid impact, which can lead to wear or failure of mechanical parts. Therefore, timely monitoring of the operating status of pump units and predicting failures can not only extend the service life of the equipment, but also effectively avoid economic losses and safety hazards caused by downtime due to failures.

[0003] At present, the existing monitoring methods of pump units mainly rely on vibration sensors or pressure sensors. These methods usually judge the status of the equipment by detecting the vibration or fluid pressure changes during the operation of the equipment. However, such methods have the following limitations: on the one hand, the installation position and number of sensors have a great influence on the monitoring accuracy, and in an environment where multiple devices are running, it is difficult to accurately distinguish the status of each device; on the other hand, these methods often cannot reflect the detailed characteristics of the mechanical faults inside the pump unit in real time, especially in a complex acoustic environment, and cannot effectively use the acoustic information of the pump unit for fault identification.

[0004] Therefore, there is an urgent need to invent a pump unit monitoring technology based on voiceprint to solve the problem that the existing pump unit monitoring method is limited by the sensor installation location and environmental complexity, and it is difficult to accurately distinguish the status of multiple devices. Summary of the invention

[0005] In view of this, the present invention proposes a fault soundprint monitoring method and system for a pump unit monitoring system, aiming to solve the problem that the existing pump unit monitoring method is limited by the sensor installation position and environmental complexity, and it is difficult to accurately distinguish the status of multiple devices.

[0006] The present invention proposes a fault soundprint monitoring method for a pump unit monitoring system, comprising: Setting a sound collection device, and acquiring environmental sounds in the monitoring area based on the sound collection device; Acquire the sound propagation path of each pump unit in the monitoring area, and determine the spatial position between the sound collection device and each pump unit according to the sound propagation path; According to the spatial position, the operating sound of each of the pump units in the environmental sound is obtained, the voiceprint feature in the operating sound is obtained, and according to the relationship between the voiceprint feature and the preset voiceprint feature vector of the pump unit, it is determined whether the pump unit has an operating fault; When it is determined that the pump unit has an operating fault, the fault type of the pump unit is determined based on the relationship between the sound print feature and each of the fault sound print features.

[0007] Furthermore, obtaining the sound propagation path of each pump unit in the monitoring area includes: Acquire the structural layout in the monitoring area, acquire the number of obstacles between each pump unit and the sound collection device and the material of each obstacle according to the structural layout, and determine the propagation loss from the pump unit to the sound collection device according to the relationship between the number of obstacles and the material of each obstacle; ; Wherein, L is the propagation loss from the pump unit to the sound collection device, L 0 is the free space propagation loss, n is the number of obstacles between the pump unit and the sound collection device, α i is the loss of sound wave absorption caused by the ith obstacle, β i is the additional loss caused by the i-th obstacle reflecting or scattering the sound wave, wherein the free space propagation loss is obtained as follows: ; Wherein, d is the straight-line distance from the pump unit to the sound collection device, f is the sound wave frequency, and C is environmental data, which includes environmental temperature and environmental humidity; The refraction angle or reflection angle between the pump unit and the sound collecting device, and between the pump unit and each obstacle are determined according to the structural layout, and the sound propagation path between the pump unit and the sound collecting device is determined according to the relationship between the propagation loss and the refraction angle or reflection angle between each obstacle.

[0008] Furthermore, when determining the spatial position between the sound collecting device and each of the pump units according to the sound propagation path, it includes: Acquire the propagation direction and propagation length of the sound propagation path, and determine the relative position between the pump unit and the sound collection device according to the propagation direction and propagation length and the structural layout; Acquire the sound collection time between each pump unit collected by the sound collection device, and acquire the time difference between each sound collection time; The coordinates of the sound collection device are determined as the center point, and the spatial position of each pump unit is determined according to the relationship between the center point, the relative position and the time difference between each sound collection time:

[0009] Among them, (x 0 ,y 0 ,z 0 ) is the spatial coordinate of the sound collection device, (x i ,y i ,z i ) is the pump unit M i The spatial coordinates of (x j ,y j ,z j ) is the pump unit M j The spatial coordinates, v is the speed of sound waves in the air, t i The sound collecting device collects the sound of the pump unit M i The sound collection time, t j The sound collecting device collects the sound of the pump unit M j The sound collection time.

[0010] Further, according to the spatial position, when obtaining and determining the operating sound of each pump unit in the environmental sound, it includes: Determine the coordinates of the sound collection device as the center point, and divide the ambient sound into several directional sound sources based on the spatial position between each of the pump units and the sound collection device; Determine the regional sound of each pump unit in each sound source according to the time difference between the sound collection times and the sound propagation path; The frequency spectrum features in the sounds of the respective regions are extracted based on Fourier transform, and the sound signal of the pump unit is determined according to the relationship between the frequency spectrum features and the preset frequency spectrum features.

[0011] Further, when determining the regional sound of each pump unit in each sound source according to the time difference between the sound collection times and the sound propagation path, it includes: Constructing a time window according to the time difference between the sound collection times in the sound source; The ambient sound in the sound source within the corresponding time is acquired according to the time window, and the regional sound of each pump unit is determined according to the sound propagation path of the ambient sound in the sound source.

[0012] Further, when determining whether the pump unit has an operation fault according to the relationship between the voiceprint feature and the preset voiceprint feature vector of the pump unit, it includes: Preprocessing the operating sound of the pump unit and extracting the voiceprint features of the preprocessed operating sound; Determine whether the pump unit has an operation fault according to the relationship between the voiceprint feature and the preset voiceprint feature vector: When the voiceprint feature is consistent with the preset voiceprint feature vector, it is determined that there is no operating fault in the pump unit; When the voiceprint feature is inconsistent with the preset voiceprint feature vector, it is determined that the pump unit has an operating fault.

[0013] Furthermore, the preprocessing of the operating sound of the pump unit includes: According to the relationship between the propagation loss of the pump unit and the pre-configured first preset propagation loss and second preset propagation loss, a correction coefficient is determined, and the amplitude and frequency characteristics of the operating sound are calibrated according to the correction coefficient: When the propagation loss is lower than the first preset propagation loss, determining the correction coefficient to be T1; When the propagation loss is higher than or equal to the first preset propagation loss, and the propagation loss is lower than the second preset propagation loss, determining the correction coefficient to be T2; When the propagation loss is higher than or equal to the second preset propagation loss, determining the correction coefficient to be T3; The first preset propagation loss is smaller than the second preset propagation loss, and T1<T2<T3<1.

[0014] Further, according to the relationship between the voiceprint feature and each of the fault voiceprint features, determining the fault type of the pump unit includes: A fault voiceprint feature set is pre-established, and matching is performed between the voiceprint feature and each of the fault voiceprint feature vectors in the fault voiceprint feature set, wherein: The fault voiceprint feature vector that is consistent with the voiceprint feature is obtained, and the fault type corresponding to the fault voiceprint feature vector is determined as the fault type of the pump unit.

[0015] Furthermore, when the fault voiceprint feature set is pre-established, it includes: Acquire each historical fault data of the pump unit, wherein the fault data includes the fault type, the fault severity value and the ambient temperature at the time of the fault; Establishing a fault correlation equation according to the historical fault data, and obtaining a distance metric between each of the fault correlation equations; Establishing a distance matrix according to the distance metric, and iteratively clustering the fault association equations according to the distance matrix; Acquire each fault association formula after iterative clustering, determine the fault type according to each fault association formula after iterative clustering, and extract the fault voiceprint feature vector of each fault association formula according to each fault association formula after iterative clustering; The fault voiceprint feature set is established according to each of the fault voiceprint feature vectors.

[0016] Compared with the prior art, the beneficial effect of the present invention is that by setting a sound collection device in the monitoring area and analyzing the sound wave propagation path, the problem of the traditional monitoring method's dependence on vibration sensors or pressure sensors is effectively overcome. In the traditional method, the installation position and number of sensors directly affect the coverage and accuracy of the monitoring, while the technical solution realizes non-contact multi-device status monitoring through the analysis of the sound propagation path, eliminates the limitations of sensor layout, and greatly improves the flexibility and applicability of the monitoring system. In addition, by combining the sound wave propagation path with the structural layout in the monitoring area, the spatial position between the sound collection device and each pump unit can be accurately determined. This spatial positioning method can not only effectively separate the operating sounds of multiple pump units, but also eliminate the influence of environmental noise and interference sound sources in a complex acoustic environment, thereby providing high-quality sound data for subsequent soundprint feature extraction, ensuring the accuracy and reliability of monitoring. Furthermore, by extracting the soundprint features in the operating sound and comparing them with the preset soundprint feature vector, it is possible to quickly determine whether the operating state of the pump unit is abnormal. As the feature vector of the operating sound of the pump unit, the soundprint feature can reflect the mechanical state changes and fault feature information of the equipment operation. Compared with fault diagnosis technology based on vibration or pressure signals, this method can capture small abnormal changes in equipment operation more comprehensively and sensitively, thereby improving the sensitivity and accuracy of fault diagnosis. Finally, when it is determined that the pump unit has an operating fault, this technical solution can further determine the fault type through the relationship between the voiceprint features and the voiceprint features of various faults. This fault type identification method based on voiceprint features can not only accurately identify faults, but also provide detailed fault feature information, which is convenient for subsequent repair and maintenance work arrangements, thereby reducing equipment downtime and maintenance costs, and improving the reliability and economic benefits of system operation.

[0017] On the other hand, the present application also provides a fault soundprint monitoring system for a pump unit monitoring system, comprising: A sound collection device is configured to obtain environmental sounds in the monitoring area; an acquisition module, electrically connected to the sound collection device, the acquisition module being configured to acquire the sound propagation path of each pump unit in the monitoring area, and determine the spatial position between the sound collection device and each pump unit according to the sound propagation path; a monitoring module, electrically connected to the acquisition module, the monitoring module being configured to acquire and determine the operating sound of each of the pump units in the environmental sound according to the spatial position, acquire the voiceprint features in the operating sound, and determine whether the pump unit has an operating fault according to the relationship between the voiceprint features and the preset voiceprint feature vector of the pump unit; The judgment module is electrically connected to the monitoring module, and is configured to determine the fault type of the pump unit according to the relationship between the voice print feature and each of the fault voice print features when it is determined that the pump unit has an operating fault.

[0018] It can be understood that the fault soundprint monitoring method and system for a pump unit monitoring system in the above-mentioned embodiments of the present invention have the same beneficial effects and will not be described in detail. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings: Figure 1 A flowchart of a fault soundprint monitoring method for a pump unit monitoring system provided by an embodiment of the present invention; Figure 2 A functional block diagram of a fault soundprint monitoring system for a pump unit monitoring system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, in the absence of conflict, the embodiments of the present invention and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0021] like Figure 1 In some embodiments of the present application, this embodiment provides a fault soundprint monitoring method for a pump unit monitoring system, including: Step S100: Setting a sound collection device, and acquiring environmental sounds in a monitoring area based on the sound collection device.

[0022] Step S200: Acquire the sound propagation path of each pump unit in the monitoring area, and determine the spatial position between the sound collection device and each pump unit according to the sound propagation path.

[0023] Specifically, when obtaining the sound propagation path of each pump unit in the monitoring area, it includes: obtaining the structural layout in the monitoring area, obtaining the number of obstacles between each pump unit and the sound collection device and the material of each obstacle based on the structural layout, and determining the propagation loss from the pump unit to the sound collection device based on the relationship between the number of obstacles and the material of each obstacle.

[0024] .

[0025] Where L is the propagation loss from the pump unit to the sound collection device, L 0 is the free space propagation loss, n is the number of obstacles between the pump unit and the sound collection device, α i is the loss of sound wave absorption caused by the ith obstacle, β i is the additional loss caused by the reflection or scattering of the sound wave by the ith obstacle, where the free space propagation loss is obtained as follows: .

[0026] Wherein, d is the straight-line distance from the pump unit to the sound collection device, f is the sound wave frequency, and C is the environmental data, which includes the ambient temperature and humidity. The refraction angle or reflection angle between the pump unit and the sound collection device, and between the pump unit and each obstacle are determined according to the structural layout, and the sound propagation path between the sound collection devices of the pump unit is determined according to the relationship between the propagation loss and the refraction angle or reflection angle between each obstacle.

[0027] Specifically, the loss of sound wave absorption caused by obstacles is obtained as follows: ; where μ i is the sound absorption coefficient of the obstacle material, t i is the thickness of the obstacle material.

[0028] Specifically, the additional loss caused by obstacles in reflecting or scattering sound waves is obtained as follows: ; Among them, ρi is the surface roughness of the obstacle, κ is the reflection loss coefficient, which is related to the sound wave frequency, and θi is the incident angle of the sound wave on the obstacle surface.

[0029] Specifically, when determining the spatial position between the sound collection device and each pump unit based on the sound propagation path, it includes: obtaining the propagation direction and propagation length of the sound propagation path, and determining the relative position between the pump unit and the sound collection device based on the propagation direction and propagation length and the structural layout. Obtaining the sound collection time between each pump unit collected by the sound collection device, and obtaining the time difference between each sound collection time. Determine the coordinates of the sound collection device as the center point, and determine the spatial position of each pump unit based on the relationship between the center point, the relative position, and the time difference between each sound collection time:

[0030] Among them, (x 0 ,y 0 ,z 0 ) is the spatial coordinate of the sound collection device, (x i ,y i ,z i ) is the pump unit M i The spatial coordinates of (x j ,y j ,z j ) is the pump unit M j The spatial coordinates, v is the speed of sound waves in the air, t i The sound collection device collects the pump unit M i The sound collection time, t j The sound collection device collects the pump unit M j The sound collection time.

[0031] It can be understood that by obtaining the structural layout information of the monitoring area, the number, material and relative spatial position of obstacles between the pump unit and the sound collection device can be clarified. The presence of obstacles will cause absorption, reflection or scattering effects on sound wave propagation, thereby affecting the propagation path and propagation loss. Using the structural layout information, the sound wave propagation path between the pump unit and the sound collection device can be established, providing a basic basis for subsequent spatial positioning and feature extraction. Secondly, according to the direction, length and relationship of the propagation path with the structural layout, the relative position between the pump unit and the sound collection device is determined. Combined with the time difference (TDOA) involved in the sound propagation path, the spatial position of the pump unit can be further inferred using the sound wave propagation velocity v and the time difference relationship. This method combines path loss, spatial direction and time information to provide a high-precision acoustic positioning method. Finally, the positioning relationship of multiple pump units can be combined to accurately determine the spatial coordinates of all pump units in the monitoring area. This formula combines the coordinates of the center point of the sound collection device, the relative position of the pump unit, the sound propagation velocity and the acquisition time difference, providing theoretical support for accurate positioning in complex acoustic environments.

[0032] It can be seen that by analyzing the structural layout in the monitoring area, including the number and material of obstacles, the propagation loss factors in the sound wave propagation path can be accurately determined. Based on the accurate calculation of the propagation path, it can not only effectively reduce the uncertainty in the sound wave propagation, but also provide reliable data support for subsequent positioning and signal analysis. Secondly, by considering the material, absorption characteristics, reflection characteristics and spatial distribution of obstacles, combined with factors such as sound wave frequency and propagation angle, the impact of obstacles on sound waves is quantified. Even in a complex industrial environment, it can adapt to different obstacle characteristics through detailed modeling of propagation loss, so as to achieve accurate sound propagation analysis and equipment positioning. At the same time, by obtaining the propagation direction and propagation length of the sound propagation path, and combining the time difference between the acquisition device and the pump unit, the spatial position of the pump unit can be accurately calculated using the center point coordinates and relative position relationship. This multi-dimensional data fusion method can significantly improve the positioning accuracy and provide accurate basic data for equipment status monitoring and fault diagnosis. Finally, by integrating propagation loss, time difference and structural layout information, the hardware cost and installation complexity are effectively reduced. The impact of misjudgment and multipath interference is also reduced through accurate propagation path analysis.

[0033] Step S300: Obtain the operating sound of each pump unit in the ambient sound according to the spatial position, obtain the voiceprint features in the operating sound, and determine whether the pump unit has an operating fault according to the relationship between the voiceprint features and the preset voiceprint feature vector of the pump unit.

[0034] Specifically, when obtaining and determining the operating sound of each pump unit in the ambient sound according to the spatial position, it includes: determining the coordinates of the sound collection device as the center point, and dividing the ambient sound into several directional sound sources based on the spatial position between each pump unit and the sound collection device. According to the time difference between the sound collection time and the sound propagation path, the regional sound of each pump unit in each sound source is determined. Based on Fourier transform, each spectrum feature in the sound of each region is extracted, and according to the relationship between each spectrum feature and the preset spectrum feature, the sound signal of the pump unit is determined.

[0035] Specifically, when determining the regional sound of each pump unit in each sound source according to the time difference between the sound collection times and the sound propagation path, it includes: constructing a time window according to the time difference between the sound collection times in the sound source. Acquiring the ambient sound in the sound source within the corresponding time according to the time window, and determining the regional sound of each pump unit according to the sound propagation path of the ambient sound in the sound source.

[0036] Specifically, when determining whether the pump unit has an operating fault according to the relationship between the voiceprint feature and the preset voiceprint feature vector of the pump unit, the method includes: preprocessing the operating sound of the pump unit and extracting the voiceprint feature of the preprocessed operating sound. When the voiceprint feature is consistent with the preset voiceprint feature vector, it is determined that the pump unit does not have an operating fault. When the voiceprint feature is inconsistent with the preset voiceprint feature vector, it is determined that the pump unit has an operating fault.

[0037] Specifically, when the operating sound of the pump unit is preprocessed, it includes: determining a correction coefficient according to the relationship between the propagation loss of the pump unit and the pre-configured first preset propagation loss and second preset propagation loss, and calibrating the amplitude and frequency characteristics of the operating sound according to the correction coefficient: when the propagation loss is lower than the first preset propagation loss, the correction coefficient is determined to be T1. When the propagation loss is higher than or equal to the first preset propagation loss, and the propagation loss is lower than the second preset propagation loss, the correction coefficient is determined to be T2. When the propagation loss is higher than or equal to the second preset propagation loss, the correction coefficient is determined to be T3. Among them, the first preset propagation loss is less than the second preset propagation loss, and T1<T2<T3<1.

[0038] It can be understood that by setting the coordinates of the sound collection device as the center point and dividing the ambient sound in the monitoring area into sound sources in several directions according to the spatial position relationship between the pump unit and the sound collection device. In this way, the sound wave sources in different directions can be identified from the ambient sound, which provides a basis for the subsequent sound source identification and positioning. By determining the sound source area where each pump unit is located, the accuracy of sound source identification can be improved. Secondly, by using the time difference between the sound collection time and the sound wave propagation path, the regional sound of each pump unit in each sound source can be further determined. By establishing a time window based on the time difference, the sound signal is divided and extracted in chronological order, so that the sound signal characteristics generated at different times can be captured. Combined with the propagation path, the sound signal of each pump unit can be accurately identified and separated from the surrounding environmental noise, reducing interference and improving the accuracy of signal processing. In addition, for the identification and analysis of the operating sound of each pump unit, the sound signal is spectrally analyzed based on Fourier transform to extract the spectral features in the sound of each area. Fourier transform can convert sound signals in the time domain into the frequency domain, thereby revealing the frequency components of the sound signal, which is crucial for distinguishing the sound signals of different pump units from complex environments. By comparing the extracted spectral features with the preset standard spectral features, the working status and fault conditions of the pump unit can be further judged. Finally, in the preprocessing of the operating sound, the voiceprint features of the sound signal are extracted. The voiceprint features are unique identifiers that describe the operating status of the equipment. By comparing the voiceprint features with the preset voiceprint feature vectors, it can be determined whether the pump unit is in normal working condition. If the voiceprint features are consistent with the preset feature vectors, it indicates that the equipment is operating normally; if they are inconsistent, it indicates that the equipment may have a fault. This method realizes automatic fault identification through feature matching, with high accuracy and real-time performance.

[0039] It can be seen that through accurate spatial positioning and segmentation technology, the ambient sound of the monitoring area is divided into sound sources in multiple directions, so that the operating sound of each pump unit can be clearly distinguished from other noise sources. This method not only improves the accuracy of sound collection, but also effectively avoids the signal interference problem when multiple pump units are running at the same time, thereby achieving more accurate sound source identification. Secondly, by combining the difference in sound collection time and the sound wave propagation path, the regional sound of each pump unit can be accurately determined. This method can construct a time window and calculate based on the propagation path of the ambient sound, thereby improving the accuracy of sound source positioning. Even in a complex environment, the exact location of each pump unit can be effectively determined, and the error caused by factors such as obstacles can be reduced, thereby enhancing the positioning effect. In addition, the frequency spectrum characteristics of the sound in each area can be extracted through Fourier transform, and the frequency composition of the running sound of the pump unit can be analyzed in detail. By comparing these spectral characteristics with the preset standard spectrum, the operating status of the pump unit can be effectively identified, and the normal operation and fault status can be distinguished. This spectrum analysis method helps to accurately identify the sound signals of the pump unit under different environmental noise backgrounds and improves the robustness of fault detection. Finally, by extracting the voiceprint features of the pump unit's operating sound and comparing them with the preset voiceprint feature vector, automatic fault detection can be achieved. When the voiceprint features are consistent, it indicates that the pump unit is operating normally; when the voiceprint features are inconsistent, the existence of a fault can be determined in a timely manner. This fault detection method is not only accurate and efficient, but can also monitor the equipment status in real time without interfering with the equipment's operation, reduce manual intervention, and improve detection efficiency.

[0040] Step S400: When it is determined that the pump unit has an operating fault, the fault type of the pump unit is determined according to the relationship between the voice print feature and each fault voice print feature.

[0041] Specifically, when determining the fault type of the pump unit based on the relationship between the voiceprint feature and each fault voiceprint feature, it includes: pre-establishing a fault voiceprint feature set, matching the voiceprint feature with each fault voiceprint feature vector in the fault voiceprint feature set, wherein: obtaining a fault voiceprint feature vector that is consistent with the voiceprint feature, and determining the fault type corresponding to the fault voiceprint feature vector as the fault type of the pump unit.

[0042] Specifically, when establishing a fault voiceprint feature set in advance, it includes: obtaining historical fault data of the pump unit, wherein the fault data includes the fault type, fault severity value and ambient temperature at the time of the fault. Establishing a fault association formula based on the historical fault data, and obtaining the distance measurement between each fault association formula. Establishing a distance matrix based on the distance measurement, and iteratively clustering each fault association formula based on the distance matrix. Obtaining each fault association formula after iterative clustering, determining the fault type according to each fault association formula after iterative clustering, and extracting the fault voiceprint feature vector of each fault association formula according to each fault association formula after iterative clustering. Establishing a fault voiceprint feature set based on each fault voiceprint feature vector.

[0043] It can be understood that by pre-establishing a fault soundprint feature set, the fault type of the pump unit is identified by using the relationship between the soundprint feature and each fault soundprint feature vector. Specifically, by collecting and analyzing the running sound of the pump unit, it is compared with the soundprint features of the historical faults. When the current soundprint feature is consistent with a feature in the fault soundprint feature set, the fault type of the pump unit can be accurately determined. This soundprint feature matching method uses the uniqueness of the acoustic signal to achieve accurate diagnosis of equipment faults. Secondly, in order to improve the accuracy of fault diagnosis, a fault correlation formula is established by obtaining the historical fault data of the pump unit (including fault type, severity value and ambient temperature, etc.). The principle of this process is to construct a fault model by analyzing the relationship between each fault feature in the historical fault data to describe the similarities and differences between different faults. This provides important basic data for subsequent fault detection, so that the diagnosis of new faults can rely on historical experience and data. In addition, after the fault correlation formula is established, the similarity between different faults is analyzed by calculating the distance metric between each fault correlation formula. The distance metric is used to quantify the similarity of fault features and help identify the type of fault that may occur under specific conditions. Through the distance matrix, iterative clustering can be further performed to classify and group historical faults according to the similarity of fault features. This clustering process can more effectively reveal and organize fault types and provide strong support for fault diagnosis. Iterative clustering is to optimize the classification and determination of fault types through repeated iterations. The principle of this process is to dynamically adjust based on the similarity between data, classify similar fault types into one category, and thus optimize the accuracy and flexibility of fault identification. After each iteration, the clustering results will be adjusted, and the final fault type will be determined based on the new clustering results, so that fault diagnosis can be gradually refined over time and improve accuracy. Finally, through the fault association formula obtained after iterative clustering, the voiceprint feature vector corresponding to each fault type is extracted, and a complete fault voiceprint feature set is established. These feature sets provide detailed reference data for voiceprint matching. By continuously updating and optimizing these feature sets, new types of faults can be adapted, making fault diagnosis more accurate. The core of this establishment process is to convert multi-dimensional historical data into effective voiceprint features, so that the fault identification of pump units can be more comprehensive, flexible and efficient.

[0044] It can be seen that, through the pre-established fault soundprint feature set, the running sound of the pump unit can be accurately compared with the historical fault soundprint features, so as to effectively identify the fault type of the pump unit. This method makes fault diagnosis more targeted and accurate by matching the relationship between the soundprint features and the soundprint features of each fault, so that different types of faults can be discovered and classified in time. Secondly, by using the historical fault data of the pump unit, including the fault type, severity and environmental factors, a more accurate prediction of possible future faults can be made by establishing a fault correlation. This data-driven method not only improves the foresight of fault detection, but also effectively converts historical data into a diagnostic basis that can be used for real-time monitoring. In addition, through iterative clustering technology, fault data can be grouped more carefully to identify the similarity and potential correlation of faults. This method effectively reduces the ambiguity of fault types, making it possible to accurately distinguish different types of faults, especially in complex environments affected by multiple factors, greatly improving the accuracy and sensitivity of fault diagnosis. At the same time, based on the matching of soundprint features with fault soundprint features, sound patterns consistent with historical fault features can be quickly identified based on the sound signals collected in real time. In this way, fault detection and fault type determination can be performed automatically without human intervention, which greatly improves monitoring efficiency and reduces the complexity of manual operation. Finally, by combining the matching of voiceprint features with cluster analysis of historical fault data, it is possible to respond quickly and accurately to pump unit faults and provide timely feedback on the fault type. This rapid response mechanism provides maintenance personnel with important fault information, helping them to perform more targeted maintenance work, thereby avoiding blind maintenance and excessive equipment downtime, and improving the operational efficiency and reliability of the equipment.

[0045] In the above embodiment, by setting a sound collection device in the monitoring area and analyzing the sound wave propagation path, the problem of the traditional monitoring method's dependence on vibration sensors or pressure sensors is effectively overcome. In the traditional method, the installation position and number of sensors directly affect the coverage and accuracy of the monitoring, while the technical solution realizes non-contact multi-device status monitoring through the analysis of the sound propagation path, eliminates the limitations of sensor layout, and greatly improves the flexibility and applicability of the monitoring system. In addition, by combining the sound wave propagation path with the structural layout in the monitoring area, the spatial position between the sound collection device and each pump unit can be accurately determined. This spatial positioning method can not only effectively separate the operating sounds of multiple pump units, but also eliminate the influence of environmental noise and interference sound sources in a complex acoustic environment, thereby providing high-quality sound data for subsequent soundprint feature extraction, ensuring the accuracy and reliability of monitoring. Furthermore, by extracting the soundprint features in the operating sound and comparing them with the preset soundprint feature vector, it is possible to quickly determine whether the operating state of the pump unit is abnormal. As the feature vector of the operating sound of the pump unit, the soundprint feature can reflect the mechanical state changes and fault feature information of the equipment operation. Compared with fault diagnosis technology based on vibration or pressure signals, this method can capture small abnormal changes in equipment operation more comprehensively and sensitively, thereby improving the sensitivity and accuracy of fault diagnosis. Finally, when it is determined that the pump unit has an operating fault, this technical solution can further determine the fault type through the relationship between the voiceprint features and the voiceprint features of various faults. This fault type identification method based on voiceprint features can not only accurately identify faults, but also provide detailed fault feature information, which is convenient for subsequent repair and maintenance work arrangements, thereby reducing equipment downtime and maintenance costs, and improving the reliability and economic benefits of system operation.

[0046] In another preferred embodiment based on the above embodiment, Figure 2 As shown, this embodiment provides a fault soundprint monitoring system for a pump unit monitoring system, including: a sound collection device, an acquisition module, a monitoring module and a judgment module.

[0047] Specifically, the sound collection device is configured to obtain environmental sounds in the monitoring area. The acquisition module is electrically connected to the sound collection device, and the acquisition module is configured to obtain the sound propagation path of each pump unit in the monitoring area, and determine the spatial position between the sound collection device and each pump unit according to the sound propagation path. The monitoring module is electrically connected to the acquisition module, and the monitoring module is configured to obtain the operating sound of each pump unit in the environmental sound according to the spatial position, obtain the voiceprint features in the operating sound, and determine whether the pump unit has an operating fault according to the relationship between the voiceprint features and the preset voiceprint feature vectors of the pump unit. The judgment module is electrically connected to the monitoring module, and the judgment module is configured to determine the fault type of the pump unit according to the relationship between the voiceprint features and the voiceprint features of each fault when it is determined that the pump unit has an operating fault.

[0048] It can be understood that the fault soundprint monitoring method and system for a pump unit monitoring system in the above-mentioned embodiments of the present invention have the same beneficial effects and will not be described in detail.

[0049] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0050] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0051] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0052] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0053] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A fault soundprint monitoring method for a pump unit monitoring system, characterized in that: include: Setting a sound collection device, and acquiring environmental sounds in the monitoring area based on the sound collection device; Acquire the sound propagation path of each pump unit in the monitoring area, and determine the spatial position between the sound collection device and each pump unit according to the sound propagation path; According to the spatial position, the operating sound of each of the pump units in the environmental sound is obtained, the voiceprint feature in the operating sound is obtained, and according to the relationship between the voiceprint feature and the preset voiceprint feature vector of the pump unit, it is determined whether the pump unit has an operating fault; When it is determined that the pump unit has an operating fault, the fault type of the pump unit is determined based on the relationship between the sound print feature and each fault sound print feature.

2. The fault soundprint monitoring method for a pump unit monitoring system according to claim 1, characterized in that: When obtaining the sound propagation path of each pump unit in the monitoring area, it includes: Acquire the structural layout in the monitoring area, acquire the number of obstacles between each pump unit and the sound collection device and the material of each obstacle according to the structural layout, and determine the propagation loss from the pump unit to the sound collection device according to the relationship between the number of obstacles and the material of each obstacle; ； Wherein, L is the propagation loss from the pump unit to the sound collection device, L0 is the free space propagation loss, n is the number of obstacles between the pump unit and the sound collection device, α i is the loss of sound wave absorption caused by the ith obstacle, β i is the additional loss caused by the i-th obstacle reflecting or scattering the sound wave, wherein the free space propagation loss is obtained as follows: ； Wherein, d is the straight-line distance from the pump unit to the sound collection device, f is the sound wave frequency, and C is environmental data, which includes environmental temperature and environmental humidity; The refraction angle or reflection angle between the pump unit and the sound collecting device, and between the pump unit and each obstacle are determined according to the structural layout, and the sound propagation path between the pump unit and the sound collecting device is determined according to the relationship between the propagation loss and the refraction angle or reflection angle between each obstacle.

3. The fault soundprint monitoring method for a pump unit monitoring system according to claim 2, characterized in that: When determining the spatial position between the sound collecting device and each of the pump units according to the sound propagation path, it includes: Acquire the propagation direction and propagation length of the sound propagation path, and determine the relative position between the pump unit and the sound collection device according to the propagation direction and propagation length and the structural layout; Acquire the sound collection time between each pump unit collected by the sound collection device, and acquire the time difference between each sound collection time; The coordinates of the sound collection device are determined as the center point, and the spatial position of each pump unit is determined according to the relationship between the center point, the relative position and the time difference between each sound collection time: ； Among them, (x0, y0, z0) is the spatial coordinate of the sound collection device, (x i ,y i ,z i ) is the pump unit M i The spatial coordinates of (x j ,y j ,z j ) is the pump unit M j The spatial coordinates, v is the speed of sound waves in the air, t i The sound collecting device collects the sound of the pump unit M i The sound collection time, t j The sound collecting device collects the sound of the pump unit M j The sound collection time.

4. The fault soundprint monitoring method for a pump unit monitoring system according to claim 3, characterized in that: When obtaining and determining the operating sound of each pump unit in the environmental sound according to the spatial position, it includes: Determine the coordinates of the sound collection device as the center point, and divide the ambient sound into several directional sound sources based on the spatial position between each of the pump units and the sound collection device; Determine the regional sound of each pump unit in each sound source according to the time difference between the sound collection times and the sound propagation path; The frequency spectrum features in the sounds of the respective regions are extracted based on Fourier transform, and the sound signal of the pump unit is determined according to the relationship between the frequency spectrum features and the preset frequency spectrum features.

5. The fault soundprint monitoring method for a pump unit monitoring system according to claim 4, characterized in that: Determining the regional sound of each pump unit in each sound source according to the time difference between the sound collection times and the sound propagation path includes: Constructing a time window according to the time difference between the sound collection times in the sound source; The ambient sound in the sound source within the corresponding time is acquired according to the time window, and the regional sound of each pump unit is determined according to the sound propagation path of the ambient sound in the sound source.

6. The fault soundprint monitoring method for a pump unit monitoring system according to claim 5, characterized in that: When determining whether the pump unit has an operation fault according to the relationship between the voiceprint feature and the preset voiceprint feature vector of the pump unit, the method includes: Preprocessing the operating sound of the pump unit and extracting the voiceprint features of the preprocessed operating sound; Determine whether the pump unit has an operation fault according to the relationship between the voiceprint feature and the preset voiceprint feature vector: When the voiceprint feature is consistent with the preset voiceprint feature vector, it is determined that there is no operating fault in the pump unit; When the voiceprint feature is inconsistent with the preset voiceprint feature vector, it is determined that the pump unit has an operating fault.

7. The fault soundprint monitoring method for a pump unit monitoring system according to claim 6, characterized in that: The pre-processing of the running sound of the pump unit includes: According to the relationship between the propagation loss of the pump unit and the pre-configured first preset propagation loss and second preset propagation loss, a correction coefficient is determined, and the amplitude and frequency characteristics of the operating sound are calibrated according to the correction coefficient: When the propagation loss is lower than the first preset propagation loss, determining the correction coefficient to be T1; When the propagation loss is higher than or equal to the first preset propagation loss, and the propagation loss is lower than the second preset propagation loss, determining the correction coefficient to be T2; When the propagation loss is higher than or equal to the second preset propagation loss, determining the correction coefficient to be T3; The first preset propagation loss is smaller than the second preset propagation loss, and T1<T2<T3<1.

8. The fault soundprint monitoring method for a pump unit monitoring system according to claim 1, characterized in that: Determining the fault type of the pump unit according to the relationship between the soundprint feature and each of the fault soundprint features includes: A fault voiceprint feature set is pre-established, and matching is performed between the voiceprint feature and each of the fault voiceprint feature vectors in the fault voiceprint feature set, wherein: The fault voiceprint feature vector that is consistent with the voiceprint feature is obtained, and the fault type corresponding to the fault voiceprint feature vector is determined as the fault type of the pump unit.

9. The fault soundprint monitoring method for a pump unit monitoring system according to claim 8, characterized in that: When establishing a fault voiceprint feature set in advance, it includes: Acquire each historical fault data of the pump unit, wherein the fault data includes the fault type, the fault severity value and the ambient temperature at the time of the fault; Establishing a fault correlation equation according to the historical fault data, and obtaining a distance metric between each of the fault correlation equations; Establishing a distance matrix according to the distance metric, and iteratively clustering the fault association equations according to the distance matrix; Acquire each fault association formula after iterative clustering, determine the fault type according to each fault association formula after iterative clustering, and extract the fault voiceprint feature vector of each fault association formula according to each fault association formula after iterative clustering; The fault voiceprint feature set is established according to each of the fault voiceprint feature vectors.

10. A fault soundprint monitoring system for a pump unit monitoring system, using a fault soundprint monitoring method for a pump unit monitoring system as claimed in any one of claims 1 to 9, characterized in that: include: A sound collection device is configured to obtain environmental sounds in the monitoring area; an acquisition module, electrically connected to the sound collection device, the acquisition module being configured to acquire the sound propagation path of each pump unit in the monitoring area, and determine the spatial position between the sound collection device and each pump unit according to the sound propagation path; a monitoring module, electrically connected to the acquisition module, the monitoring module being configured to acquire and determine the operating sound of each of the pump units in the environmental sound according to the spatial position, acquire the voiceprint features in the operating sound, and determine whether the pump unit has an operating fault according to the relationship between the voiceprint features and the preset voiceprint feature vector of the pump unit; The judgment module is electrically connected to the monitoring module, and is configured to determine the fault type of the pump unit according to the relationship between the voice print feature and each of the fault voice print features when it is determined that the pump unit has an operating fault.

Images