Method and apparatus for constructing equipment noise detection model based on industrial noise analysis technology

Abstract

This application relates to a method and apparatus for constructing an equipment noise detection model based on industrial noise analysis technology, belonging to the technical field of equipment noise detection. The method includes: acquiring sound sampling data in response to a request; processing the sound sampling data to obtain sound reconstruction data; acquiring first volume information and first density information of the sound reconstruction data in each frequency band over any time period; and constructing an equipment noise detection model based on the first volume information and the first density information. This application has the effect of detecting abnormal sound frequency bands and the volume and density distribution relationships existing in each frequency band.

Description

Technical Field

[0001] This application relates to the technical field of equipment noise detection, and in particular to an intelligent method and device for equipment noise detection based on industrial noise analysis technology. Background Technology

[0002] You can identify common equipment operating conditions by listening to the sounds. This is mainly because vibrations or noises generated during equipment operation are symptoms of abnormal conditions. Therefore, you can check the condition of some equipment and identify whether a malfunction has occurred by listening to the sounds.

[0003] Traditional Fourier analysis: The calculation basis of Fourier analysis is that the sound in a certain frequency band has a good periodicity within the analysis time. However, industrial noise is generally sudden, decays rapidly, and has a very short duration with no periodicity, which is completely inconsistent with the actual environmental sound.

[0004] Traditional wavelet analysis: In noisy environments, sound signals are very complex and do not have a consistent attenuation pattern. Wavelet analysis cannot accurately analyze all sound signals with suitable wavelet operators.

[0005] Machine learning: Machine learning has poor generalization ability, requires a large amount of abnormal sample data, and the recognition effect is not ideal. Moreover, it cannot discover periodic sound patterns and cannot accurately calculate indicators such as volume density of each frequency band. Currently, it can only intelligently recognize sound signals with obvious characteristics.

[0006] Therefore, traditional analysis methods cannot accurately analyze the frequency bands of abnormal sounds and the distribution relationship of volume and density in each frequency band. Summary of the Invention

[0007] The purpose of this application is to provide a method and apparatus for constructing a device noise detection model based on industrial noise analysis technology, which can detect the frequency bands of abnormal sounds and the volume and density distribution relationship of each frequency band.

[0008] Firstly, this application provides a method for constructing a device noise detection model based on industrial noise analysis technology, which adopts the following technical solution:

[0009] A method for constructing an equipment noise detection model in industrial noise analysis technology includes:

[0010] In response to the request to obtain sound sampling data;

[0011] The sound sampling data is processed to obtain sound reconstruction data;

[0012] Obtain the first volume information and first density information of the sound in each frequency band of the sound reconstruction data at any time period;

[0013] A device noise detection model is constructed based on the first volume information and the first density information.

[0014] By adopting the above technical solution, noise detection of the equipment is carried out based on the constructed equipment noise detection model, thereby discovering the frequency bands of abnormal sounds and the volume and density distribution relationship of each frequency band.

[0015] Optionally, processing the sound sampling data to obtain sound reconstruction data specifically includes:

[0016] Divide the frequency range into multiple frequency bands within a preset frequency range;

[0017] Obtain the second volume information and second density information of the sound sampling data in each frequency band;

[0018] The sound reconstruction data is obtained based on the second volume information and the second density information.

[0019] By adopting the above technical solution, multiple frequency bands are divided, and the second volume information and second density information of the sound sampling data are collected in each frequency band. Then, the sound reconstruction data can be obtained based on the second volume information and second density information.

[0020] Optionally, obtaining the first volume information and first density information of the sound in each frequency band of the sound reconstruction data at any time period specifically includes:

[0021] Collect the attribute information of the sound in the sound reconstruction data within a preset time range;

[0022] The attribute information is analyzed to obtain the first volume information and the first density information of the sound in each frequency band within a preset frequency range.

[0023] By adopting the above technical solution, the sound attribute information in the sound reconstruction data is collected within a preset time range, and the first volume information and first density information of each frequency band in the preset frequency range can be obtained based on the sound attribute information in the sound reconstruction data. This makes it convenient to obtain the first volume information and first density information of each frequency band in the sound reconstruction data at any time period.

[0024] Optionally, the step of constructing a device noise detection model based on the first volume information and the first density information specifically includes:

[0025] The sound reconstruction data is classified according to preset rules to obtain classification results, which include periodic sound, continuous sound and sudden sound.

[0026] Based on the first volume information and the first density information, corresponding mathematical models are constructed for the periodic sound, the continuous sound, and the sudden sound, respectively.

[0027] By adopting the above technical solution, the sound reconstruction data can be classified into periodic sound, continuous sound, and sudden sound. Then, based on the first volume information and the first density information of each frequency band, mathematical models of various sounds can be constructed respectively, so as to judge the collected sound based on the mathematical models of various sounds.

[0028] Optionally, the mathematical model of the periodic sound includes:

[0029] P_S = (T,f,v,s)

[0030] Where P_S represents the periodic sound, T represents the sound period time, f represents the sound frequency, v represents the volume, and s represents the periodic stability.

[0031] Optionally, the mathematical model of the continuous sound includes:

[0032] L_S=(f1, f2, T, t0, v, d)

[0033] Where L_S represents continuous sound, f1 and f2 represent sound frequencies, T represents the sound cycle time, t0 represents the extraction interval time, v represents volume, and d represents density.

[0034] Optionally, the mathematical model of the sudden sound includes:

[0035] S_S=(f1, f2, v, d, T, t0, v0, d0)

[0036] Where S_S represents the sudden sound, f1 and f2 represent the sound frequencies, v represents the volume, d represents the density, T represents the sound cycle time, t0 represents the extraction interval time, v0 represents the volume threshold, and d0 represents the density threshold.

[0037] Secondly, the equipment noise detection model framework device based on industrial noise analysis technology provided in this application adopts the following technical solution:

[0038] A device for constructing equipment noise detection models based on industrial noise analysis technology, comprising:

[0039] The first acquisition module is used to acquire sound sampling data in response to a request;

[0040] The data processing module is used to process the sound sampling data to obtain sound reconstruction data;

[0041] The second acquisition module is used to acquire the first volume information and the first density information of the sound in each frequency band of the sound reconstruction data at any time period;

[0042] The model building module is used to build a noise detection model based on the first volume information and the first density information.

[0043] By adopting the above technical solution, noise detection of the equipment is carried out based on the constructed equipment noise detection model, thereby discovering the frequency bands of abnormal sounds and the volume and density distribution relationship of each frequency band.

[0044] Thirdly, the terminal provided in this application adopts the following technical solution:

[0045] A terminal includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein when the processor loads the computer program, it executes the method of the first aspect.

[0046] By adopting the above technical solution, the method of the first aspect generates a computer program and stores it in the memory so that it can be loaded and executed by the processor. Thus, the user can establish a connection with the device through the terminal and query the various contents processed by the device.

[0047] Fourthly, the computer-readable storage medium provided in this application adopts the following technical solution:

[0048] A computer-readable storage medium storing a computer program that, when loaded by a processor, executes the method of the first aspect.

[0049] By adopting the above technical solution, the method of the first aspect is used to generate a computer program and store it in a computer-readable storage medium. Once the computer-readable storage medium is loaded into any computer, the method of the first aspect can be executed. Attached Figure Description

[0050] Figure 1 This is a flowchart of steps S100-S400 in an embodiment of this application;

[0051] Figure 2 This is a flowchart of steps S210-S230 in an embodiment of this application;

[0052] Figure 3 This is a schematic diagram of the volume frequency curve and density frequency curve of an embodiment of this application;

[0053] Figure 4 This is a flowchart of steps S310-S320 in an embodiment of this application;

[0054] Figure 5 This is a flowchart of steps S410-S420 in an embodiment of this application;

[0055] Figure 6 This is a flowchart illustrating the sound data acquisition process of the device according to an embodiment of this application;

[0056] Figure 7 This is a functional layer schematic diagram of the device according to an embodiment of this application. Detailed Implementation

[0057] The following is in conjunction with the appendix Figure 1 - Appendix Figure 7 This application will be described in further detail below.

[0058] To date, the instruments used both domestically and internationally for equipment fault diagnosis or product quality inspection based on sound detection fall into two categories: First, vibration meters. Vibration detection can only obtain the corresponding amplitude when equipment experiences intermittent collisions, thus objectively exhibiting low sensitivity and susceptibility to external interference. However, most equipment quality problems involve continuous squeezing friction, a phenomenon that only generates noise and does not easily induce vibration, making it difficult for vibration detection to identify problems. Second, sound spectrum analyzers. Spectrum analyzers primarily rely on traditional Fourier analysis to draw conclusions. The calculation basis of Fourier analysis is that sound in a certain frequency band has a good periodicity within the analysis time. However, actual industrial sound signals are very unstable, and Fourier analysis cannot accurately analyze data from each frequency band.

[0059] Therefore, it can be seen that these two instruments are inherently limited by their principles. They cannot detect equipment abnormalities from noisy industrial sounds and cannot completely solve the problems of equipment fault diagnosis or product quality inspection.

[0060] To address the aforementioned issues, this application provides a method for constructing a device noise detection model, thereby resolving the aforementioned technical problems.

[0061] This application discloses a method for constructing a device noise detection model based on industrial noise analysis technology, referring to... Figure 1 It includes the following steps:

[0062] S100: Responds to a request to obtain sound sampling data.

[0063] In one embodiment of this application, several devices of several models are selected, and sound sampling data is collected at the same position under the same rotation speed. For example, sound acquisition sensors are set at locations prone to failure, such as bearings, gear meshing points, and equipment. In this embodiment, a hearing diagnostic instrument is selected as the sound acquisition sensor, and then the hearing diagnostic instrument is used to collect sound sampling data of bearings, gear meshing points, and equipment.

[0064] S200: Processes the sound sampling data to obtain sound reconstruction data.

[0065] In one embodiment of this application, reference is made to Figure 2 S200 specifically includes the following steps:

[0066] S210: Divide the frequency range into multiple frequency bands within a preset frequency range.

[0067] Specifically, in this embodiment, the preset frequency range is set to 0Hz-640kHz, thereby customizing multiple frequency bands from low to high frequency within the 0Hz-640kHz range, such as 0Hz-100kHz, 100kHz-200kHz, 200kHz-300kHz, 300kHz-400kHz, 400kHz-500kHz, 500kHz-640kHz, etc.

[0068] S220: Acquire the second volume information and second density information of the sound sampling data in each frequency band.

[0069] Specifically, in this embodiment, referring to Figure 3 The second volume information includes a volume frequency curve, and the second density information includes a density frequency curve. Sound is collected by a preset volume acquisition sensor, and the software automatically constructs a volume frequency curve and a density frequency curve at a certain rotation speed based on the collected sound.

[0070] Specifically, in this embodiment, the increase in the high-frequency density curve is generally due to direct scraping or collision between metal materials, which may be caused by equipment wear, defects, pitting, deformation, unbalanced installation, or insufficient lubrication.

[0071] For example, referring to the figure, it can be shown that the density and volume of device anomalies are divided in each range. It can be seen that the frequency distribution of volume anomalies is between 6000HZ and 34000HZ, and the density anomalies are concentrated between 9000HZ and 24000HZ.

[0072] S230: Obtain sound reconstruction data based on the second volume information and the second density information.

[0073] In one embodiment of this application, sound reconstruction data can be obtained from the volume frequency curve and the density frequency curve.

[0074] S300: Acquire the first volume information and first density information of the sound in each frequency band of the sound reconstruction data at any time period.

[0075] In one embodiment of this application, reference is made to Figure 4 The S300 specifically includes the following steps:

[0076] S310: Collects sound attribute information from sound reconstruction data within a preset time range.

[0077] S320: Analyze the attribute information to obtain the first volume information and the first density information of the sound wave in each frequency band within the second preset frequency range.

[0078] S400: Construct a device noise detection model based on the first volume information and the first density information.

[0079] In one embodiment of this application, reference is made to Figure 5 The S400 specifically includes the following steps:

[0080] S410: Classify the sound reconstruction data according to preset rules to obtain classification results, and the classification results include periodic sound, continuous sound and sudden sound.

[0081] S420: Construct corresponding mathematical models for periodic sounds, continuous sounds, and sudden sounds based on the first volume information and the first density information.

[0082] Specifically, in this embodiment, the mathematical model for periodic sound is:

[0083] P_S = (T,f,v,s)

[0084] Where P_S represents the periodic sound, T represents the sound period time, f represents the sound frequency, v represents the volume, and s represents the periodic stability.

[0085] Among them, periodic sounds are captured, and the causes of equipment failures are determined by analyzing the period, frequency, volume and period stability, combined with the transmission cycle of the components of the equipment itself. A certain sound emitted at fixed intervals during the operation of the equipment is called a periodic sound.

[0086] Specifically, in this embodiment, the unit of the sound production cycle time is ms. For example, if a damaged gear makes an abnormal noise every 100ms when it rotates once, then 100ms is the sound production cycle.

[0087] Specifically, in this embodiment, the unit of sound frequency is Hz, which indicates what kind of sound is emitted. Different frequencies reflect different sound content, such as electromagnetic noise, metallic knocking, and ambient noise.

[0088] Specifically, in this embodiment, the unit of volume is dB, which represents the loudness of the emitted sound. Generally, the highest volume in the interval or the average of the highest volumes in multiple intervals is selected to characterize the feature.

[0089] Specifically, in this embodiment, the periodic stability is expressed as a percentage, representing the degree of stability of the occurrence of this period.

[0090] Specifically, in this embodiment, the mathematical model for continuous sound is:

[0091] L_S=(f1, f2, T, t0, v, d)

[0092] Where L_S represents continuous sound, f1 and f2 represent sound frequencies, T represents the sound cycle time, t0 represents the extraction interval time, v represents volume, and d represents density.

[0093] Among them, the sound frequency bands that need to be focused on are divided (or can be divided at any granularity and overlap within the entire sound frequency range), and the sound changes in these frequency bands are monitored to examine the health status of the equipment; the continuous sounds emitted during the equipment rotation process, such as friction sounds, water leakage sounds, and airflow sounds caused by the equipment rotation, are called continuous sounds.

[0094] Specifically, in this embodiment, for example, f1 is set to 5000 Hz, f2 is set to 5500 Hz, and f1-f2 is set to 5000 Hz-5500 Hz. The key audio frequency bands to be focused on can be determined by sound spectrum analysis or can be manually divided according to the operating mechanism of the device.

[0095] Specifically, in this embodiment, the maximum volume of this audio frequency band is taken every t0 time within the observation time T, thus obtaining a total of T / t0 volume data points, and the average value is taken as the volume value.

[0096] Specifically, in this embodiment, the mathematical model for the sudden sound is:

[0097] S_S=(f1, f2, v, d, T, t0, v0, d0)

[0098] Where S_S represents the sudden sound, f1 and f2 represent the sound frequencies, v represents the volume, d represents the density, T represents the sound cycle time, t0 represents the extraction interval time, v0 represents the volume threshold, and d0 represents the density threshold.

[0099] Among them, the target type of sound to be captured is set according to the sound frequency range. When sound information that meets the density and volume requirements within the sound duration is captured, i.e., the expression S is satisfied, it can be determined that a sudden sound has occurred. Such sounds that are very short in duration and do not form a periodic pattern, such as explosions, impacts, gunshots, thunder, footsteps, animal calls, and human shouts, are called sudden sounds.

[0100] Specifically, in this embodiment, for example, f1 is set to 5000 Hz, f2 is set to 5500 Hz, and f1-f2 is set to 5000 Hz-5500 Hz. The key audio frequency bands to be focused on can be determined by sound spectrum analysis or can be manually divided according to the operating mechanism of the device.

[0101] Specifically, in this embodiment, the pronunciation cycle time T is calculated by calculating the maximum value of T, and satisfying that within the time range of T, the average maximum volume of any consecutive N t0 times is greater than v0, and the average density of this frequency band is greater than d0.

[0102] In one embodiment of this application, reference is made to Figure 6 The device collects sound data through a stethoscope and uploads the collected sound data to a cloud computing cluster in real time. Then, with the help of cloud supercomputing, the target information is displayed on the display terminal.

[0103] The implementation principle of this application embodiment is as follows: sound sampling data is collected by a sound stethoscope, and then the sound sampling data is processed to obtain sound reconstruction data. Then, the first volume information and first density information of the sound reconstruction data in each frequency band at any time period are collected by a preset sensor. Finally, a device noise detection model is constructed based on the first volume information and the first density information. In this way, the frequency band of the abnormal sound and the relationship between the volume and density distribution of each frequency band can be found.

[0104] In one embodiment of this application, the application of the solution in the factory quality inspection of equipment (rotating equipment) is specifically as follows:

[0105] For each model of normal equipment, select several noise detection locations. For each noise detection location, collect several sets of noise data at different speeds of the equipment to establish a standard curve of volume and density at a certain speed.

[0106] For example, when selecting device model X, at a speed of V, select noise detection positions P1, P2, P3...Pn. Collect noise data three times at each noise detection position, with the same collection time each time, generally set to 10-15 seconds. Collect 15 sets of noise data for the same model of device. That is, each noise detection position P will generate 45 sets of noise data. For the volume and density data of each frequency band, remove the maximum value (9 sets) and the minimum value (6 sets), and take the average of the remaining 30 sets of noise data. In this way, a standard curve for the volume and density of each frequency band of device X at speed V and noise detection position P can be established. The noise standard curve can be automatically calculated and generated by software after the data is sampled and saved.

[0107] In one embodiment of this application, during testing, the device selects model X to simultaneously collect noise data at speed V and N noise detection positions. The noise data at each noise detection position is compared with the corresponding standard curve (the standard curve of model X at speed V and noise detection position P). The data of a specific frequency band is examined. When the density reaches a certain threshold (e.g., 1%), the density or volume deviation reaches a set threshold (e.g., 0.5% or 4 dB), indicating that the device has a defect. Of course, if obvious periodic sound is found at a certain noise detection position, the device may also have a defect.

[0108] In the product design and product quality inspection stages, the product's operating characteristics are evaluated using quantitative indicators based on periodic sound data and frequency band sound volume density data. For example, the sound produced by a certain product is in the low-frequency band of around 1000 Hz, the sub-frequency band is mainly airflow sound, the sound around 6000 Hz is mainly metal friction sound or echo caused by metal impact, and the frequency band above 30000 Hz is mainly the sound emitted by the moment of strong metal impact.

[0109] This system collects the sounds emitted by equipment and analyzes their operating characteristics in real time, providing a clear overview of the operational status of each part and ensuring the quality of the equipment before it leaves the factory. It can accurately identify sound changes caused by internal material issues (cracks, pores) and external problems (wear, pitting, deformation, burrs), as well as issues related to installation accuracy, loose fasteners, and failure of transmission components such as bearings. This significantly improves the equipment's factory quality inspection pass rate and fault diagnosis accuracy. By collecting the sound standard curves of equipment or products, a sound quantification mathematical model is established, eliminating the need for abnormal sound sample learning for identification. This greatly reduces labor and training costs, saves time, and improves production efficiency. Due to its high-precision identification and detection capabilities, it significantly improves the equipment's factory pass rate and reduces the equipment return rate for repair.

[0110] In one embodiment of this application, the application of the solution in equipment inspection is specifically as follows:

[0111] Longitudinal comparison: Comparison of sound curves at different times based on the same equipment and the same noise inspection location.

[0112] Specifically, in this embodiment, the same equipment is inspected, and several noise inspection points are selected on the equipment. At regular intervals, noise data is collected from these noise inspection points while the rotation speed is the same. The data is then compared with the historical sound curves of the same noise inspection point at different times to determine the aging or damage of the equipment.

[0113] Horizontal comparison: Comparison of sound curves of different devices of the same model and at the same noise inspection location.

[0114] Specifically, in this embodiment, several devices of the same model are inspected, and several noise inspection points are selected on the devices. In each cycle, under the condition of the same rotation speed, noise data is collected and compared at the same noise inspection points of different devices, so as to determine which devices have different sound curves at which noise inspection points compared with other devices, thereby providing decision support for equipment maintenance.

[0115] In one embodiment of this application, the application of the solution in online monitoring of equipment is specifically as follows:

[0116] Select several noise monitoring points on the equipment, and collect and analyze data such as the volume, density, and periodicity of sound in each frequency band of these noise monitoring points in real time. During the operation of the equipment, if the monitoring data of a certain noise monitoring point exceeds the original set threshold for a certain period of time, the noise monitoring point will be alarmed, and the fault type will be preliminarily determined based on the current sound data and historical experience data.

[0117] In one embodiment of this application, mathematical expressions for periodic sound, continuous sound, and sudden sound can be constructed to form a fully quantifiable industrial noise analysis technology model, thereby providing excellent quantitative analysis capabilities for periodic sound, continuous sound, and sudden sound. Based on the quantitative decomposition of the mathematical model, no large amount of abnormal sample data or learning is required.

[0118] Referring to the figures, in one embodiment of this application, the device detection process specifically includes: placing a hearing diagnostic instrument on the device in an adsorption and / or placement manner to collect noise data generated during device operation, uploading the collected noise data to the cloud, and then performing real-time calculation and analysis on the noise data through the cloud to obtain the calculation and analysis results, and sending the obtained calculation and analysis results to a display terminal for restriction; the display terminal in this embodiment can be set as a tablet computer, of course, depending on the actual use, the display terminal can also be set as a desktop computer or a mobile phone, this application does not limit this; and for projects with particularly strict confidentiality requirements, the computing service supports local deployment.

[0119] Compared to Fourier analysis and wavelet analysis, this method can completely quantify and construct mathematical expressions for all sounds, forming a fully quantifiable industrial noise analysis model. It has excellent quantitative analysis capabilities for periodic, continuous, and sudden sounds, and can detect anomalies in the early stages of a fault. Furthermore, since it is based on the quantitative decomposition of a mathematical model, it does not require a large amount of abnormal sample data or learning.

[0120] This application discloses a device for constructing a device noise detection model based on industrial noise analysis technology, specifically including a first acquisition module, a data processing module, a second acquisition module, and a model construction module; wherein, the first acquisition module is used to acquire sound sampling data in response to a request; the data processing module is used to process the sound sampling data to obtain sound reconstruction data; the second acquisition module is used to acquire the first volume information and the first density information of the sound in each frequency band of the sound reconstruction data in any time period; and the model construction module is used to construct a noise detection model based on the first volume information and the first density information.

[0121] In one embodiment of this application, when constructing an equipment noise detection model using an equipment noise detection model construction device based on industrial noise analysis technology, the equipment noise detection model construction method based on industrial noise analysis technology described above is used. Therefore, the specific application of the device will not be elaborated here.

[0122] In one embodiment of this application, reference is made to Figure 7 The device described in this application can be functionally divided into a data acquisition layer, a cloud supercomputing layer, and a user interaction layer. Specifically, the data acquisition layer provides data support for the real-time computing and processing of the entire platform. The data source is mainly high-frequency acquisition equipment, which has an ultra-high frequency acquisition frequency, ensuring the reliability and comprehensiveness of the acquired data. The data acquisition layer includes a parameter setting module, a main monitoring module, an alarm module, and an interface API. The cloud supercomputing layer analyzes the data, rapidly analyzes noise characteristics through a large-scale computing cluster in a supercomputing center, and transmits the results down to the front end for display, achieving millisecond-level data analysis and computation. The cloud-deployed server effectively reduces the user's deployment costs. It also supports local server deployment; the user interaction layer can include a parameter setting unit, a main monitoring unit, an alarm unit, and a data interface unit. The parameter setting unit is used to set parameters, the main monitoring unit is used to monitor the operation of the device and changes in parameters, the alarm unit is used to issue alarms when the device malfunctions or the parameters are not within a reasonable range, so as to remind relevant personnel to pay attention, and the data interface unit can connect to external electronic devices to transmit data or information in the device to external electronic devices. The device monitors noise data in real time to realize abnormal alarms. For users, it can provide data interfaces in the form of API, database docking, etc., to push data to the user system in real time.

[0123] This application discloses a terminal, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it employs the equipment noise detection model construction method based on industrial noise analysis technology described above.

[0124] In one embodiment of this application, the terminal may be a desktop computer, a laptop computer, or a cloud server, and the terminal may include, but is not limited to, a processor and a memory. For example, the terminal may also include input / output devices, network access devices, and buses.

[0125] In one embodiment of this application, the processor may be a central processing unit (CPU). Of course, depending on the actual use, other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), off-the-shelf programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. may also be used. The general-purpose processor may be a microprocessor or any conventional processor, etc., and this application does not limit it in this regard.

[0126] In one embodiment of this application, the memory can be an internal storage unit of the terminal, such as the terminal's hard disk or memory, or an external storage device of the terminal, such as a plug-in hard disk, smart memory card (SMC), secure digital card (SD), or flash memory card (FC) equipped on the terminal. The memory can also be a combination of the terminal's internal storage unit and external storage device. The memory is used to store computer programs and other programs and data required by the terminal. The memory can also be used to temporarily store data that has been output or will be output. This application does not limit this.

[0127] By configuring this terminal, the device noise detection model construction method based on industrial noise analysis technology described in the above embodiment is stored in the terminal's memory and loaded and executed on the terminal's processor. Thus, the user can establish a connection with the device through the terminal and query the various contents processed by the device.

[0128] This application discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it employs the equipment noise detection model construction method based on industrial noise analysis technology described above.

[0129] In one embodiment of this application, the computer program may be stored in a computer-readable storage medium. The computer program includes computer program code, which may be in the form of source code, object code, executable file, or certain middleware. The computer-readable storage medium includes any entity or device capable of carrying computer program code, recording medium, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the computer-readable storage medium includes, but is not limited to, the above-mentioned components.

[0130] By using this computer-readable storage medium, the equipment noise detection model construction method based on industrial noise analysis technology of the above embodiments is stored in the computer-readable storage medium and loaded and executed on the processor. After the computer-readable storage medium is loaded into any computer, any computer can execute the equipment noise detection model construction method based on industrial noise analysis technology of the above embodiments.

[0131] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Identical components are represented by the same reference numerals. Therefore, all equivalent changes made to the structure, shape, and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for constructing a device noise detection model based on industrial noise analysis technology, characterized in that, include: In response to the request to obtain sound sampling data; The sound sampling data is processed to obtain sound reconstruction data; Obtain the first volume information and first density information of the sound in each frequency band of the sound reconstruction data at any time period; A device noise detection model is constructed based on the first volume information and the first density information; The sound sampling data is processed to obtain sound reconstruction data, specifically including: Divide the frequency range into multiple frequency bands within a preset frequency range; Obtain the second volume information and second density information of the sound sampling data in each frequency band; The sound reconstruction data is obtained based on the second volume information and the second density information; The step of obtaining the first volume information and first density information of the sound in each frequency band of the sound reconstruction data at any time period specifically includes: Collect the attribute information of the sound waves in the sound reconstruction data within a preset time range; The attribute information is analyzed to obtain the first volume information and the first density information of the sound wave in each frequency band within the second preset frequency range; The step of constructing a device noise detection model based on the first volume information and the first density information specifically includes: The sound reconstruction data is classified according to preset rules to obtain classification results, which include periodic sound, continuous sound and sudden sound. Based on the first volume information and the first density information, corresponding mathematical models are constructed for the periodic sound, the continuous sound, and the sudden sound, respectively.

2. The method for constructing a device noise detection model based on industrial noise analysis technology according to claim 1, characterized in that, The mathematical model of the periodic sound includes: P_S = (T,f,v,s) Where P_S represents the periodic sound, T represents the sound period time, f represents the sound frequency, v represents the volume, and s represents the periodic stability.

3. The method for constructing an equipment noise detection model based on industrial noise analysis technology according to claim 1, characterized in that, The mathematical model for the continuous sound includes: L_S=(f1, f2, T, t0, v, d) Where L_S represents continuous sound, f1 and f2 represent sound frequencies, T represents the sound cycle time, t0 represents the extraction interval time, v represents volume, and d represents density.

4. The method for constructing an equipment noise detection model based on industrial noise analysis technology according to claim 1, characterized in that, The mathematical model for the sudden sound includes: S_S=(f1, f2, v, d, T, t0, v0, d0) Where S_S represents the sudden sound, f1 and f2 represent the sound frequency, v represents the volume, d represents the density, T represents the sound cycle time, t0 represents the extraction interval time, v0 represents the volume threshold, and d0 represents the density threshold.

5. A device for constructing a device noise detection model based on industrial noise analysis technology, characterized in that, include: The first acquisition module is used to acquire sound sampling data in response to a request; The data processing module is used to process the sound sampling data to obtain sound reconstruction data; The second acquisition module is used to acquire the first volume information and the first density information of the sound in each frequency band of the sound reconstruction data at any time period; The model building module is used to build a noise detection model based on the first volume information and the first density information; This involves processing the sound sampling data to obtain sound reconstruction data, specifically including: Divide the frequency range into multiple frequency bands within a preset frequency range; Acquire the second volume information and second density information of the sound sampling data in each frequency band; The sound reconstruction data is obtained based on the second volume information and the second density information; Obtain the first volume information and first density information of the sound in each frequency band of the sound reconstruction data at any time period, specifically including: Collect the property information of sound waves in the sound reconstruction data within a preset time range; The attribute information is analyzed to obtain the first volume information and the first density information of the sound wave in each frequency band within the second preset frequency range; A device noise detection model is constructed based on the first volume information and the first density information, specifically including: The sound reconstruction data is classified according to preset rules to obtain classification results, which include periodic sound, continuous sound and sudden sound. Based on the first volume information and the first density information, corresponding mathematical models are constructed for periodic sounds, continuous sounds, and sudden sounds.

6. A terminal, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor loads the computer program, it executes the method of any one of claims 1-4.

7. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is loaded by the processor, it executes the method of any one of claims 1-4.

Images