Self-cleaning method, device and equipment for acoustic signal acquisition device and storage medium

Abstract

The invention provides an acoustic signal acquisition device self-cleaning method, device and equipment and a storage medium. The method comprises the following steps: acquiring acoustic signals monitored by each microphone channel in an acoustic signal acquisition device; extracting a time domain feature, a frequency domain feature and a time frequency feature in the acoustic signal monitored by each microphone channel, inputting the time domain feature, the frequency domain feature and the time frequency feature of each microphone channel, the current environment data and the cleaning record into a self-cleaning start control model, and outputting a cleaning demand score of each microphone channel; according to the cleaning demand score of each microphone channel, starting self-cleaning control on the microphone channel lower than a score threshold; the self-cleaning control of the sound signal acquisition device is realized, the danger caused by manual maintenance is avoided, the self-cleaning starting opportunity can be accurately predicted, the specified channel can be independently cleaned, the resource use is minimized, and the performance and the reliability of the sound signal acquisition device are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent mining, and particularly relates to a self-cleaning method, device, equipment and storage medium for an acoustic signal acquisition device. Background Art

[0002] In underground mining operations, acoustic signal acquisition devices such as microphone arrays are crucial for acoustic monitoring and communication. However, due to the accumulation of dust and moisture, the performance of the microphone array is likely to decline. Therefore, it is necessary to clean the microphone array. The traditional method for cleaning the microphone array is to manually replace the microphone cover regularly. This method is manually operated, with low efficiency. Moreover, since the microphone array has multiple microphone channels and a large area, and the dust and moisture contaminated on each microphone channel are different, it is not easy to determine the replacement time of the microphone cover by using the method of uniformly replacing the microphone cover. In addition, due to the underground operation environment, manual operation is dangerous. Summary of the Invention

[0003] The present invention provides a self-cleaning method, device, equipment and storage medium for an acoustic signal acquisition device, so as to solve the defects that it is not easy to determine the replacement time of the microphone cover for the traditional acoustic signal acquisition device and there is danger.

[0004] The present invention provides a self-cleaning method for an acoustic signal acquisition device, including: Obtaining the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device; Extracting the time-domain features, frequency-domain features and time-frequency features from the acoustic signals monitored by each microphone channel, inputting the time-domain features, frequency-domain features and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records into the self-cleaning start control model, and outputting the cleaning requirement score of each microphone channel; Starting the self-cleaning control for the microphone channels with the cleaning requirement score lower than the score threshold according to the cleaning requirement score of each microphone channel; Wherein, the self-cleaning start control model is trained based on the historical audio signal features, historical environmental data and historical cleaning records monitored by each microphone channel.

[0005] According to the self-cleaning method for an acoustic signal acquisition device provided by the present invention, the extracting the time-domain features, frequency-domain features and time-frequency features from the acoustic signals monitored by each microphone channel includes: The time-domain features extracted from the acoustic signals monitored by each microphone channel include mean, variance, peak value and kurtosis; The frequency-domain features extracted from the acoustic signals monitored by each microphone channel include spectral peak value, spectral centroid and spectral variance; Extract the time-frequency features in the acoustic signals monitored by each microphone channel, including time-frequency feature wavelet coefficients or energy distribution diagrams.

[0006] According to the self-cleaning method of the acoustic signal acquisition device provided by the present invention, inputting the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records into the self-cleaning start control model, and outputting the cleaning requirement scores for each microphone channel, including: Fuse the time-domain features, frequency-domain features, and time-frequency features of each microphone channel to generate a multi-dimensional feature vector; Combine the current environmental data and historical cleaning records to generate a comprehensive feature matrix; Input the multi-dimensional feature vector and the comprehensive feature matrix into the trained self-cleaning start control model, and output the cleaning requirement scores for each microphone channel.

[0007] According to the self-cleaning method of the acoustic signal acquisition device provided by the present invention, the training method of the self-cleaning start control model includes: Collect the audio signal data of each microphone channel in the normal state and the polluted state, and label the cleaning requirements; Extract the time-domain features, frequency-domain features, and time-frequency domain features from the audio signals, input the time-domain features, frequency-domain features, and time-frequency domain features, as well as the historical environmental data and historical cleaning records into the self-cleaning start control model, and output the predicted cleaning requirement scores for each microphone channel; Construct a loss function according to the predicted cleaning requirement scores and the labeled cleaning requirements, adjust the parameters of the self-cleaning start control model to optimize the loss function until the training end condition is met, and obtain the trained self-cleaning start control.

[0008] According to the self-cleaning method of the acoustic signal acquisition device provided by the present invention, it further includes: Obtain the historical cleaning records and the environmental change data within a period of time; Input the historical cleaning records and the environmental change data within a period of time into the self-cleaning prediction model, and output the next self-cleaning start time; Wherein, the self-cleaning prediction model is trained based on the historical cleaning records and the environmental change data sequence.

[0009] According to the self-cleaning method of the acoustic signal acquisition device provided by the present invention, it further includes: Obtain the historical environmental data, system operation data, and fault records; Input the historical environmental data, system operation data, and fault records into the life prediction model to obtain the remaining life of the system; Output a maintenance plan based on the remaining life of the system; Among them, the life prediction model is trained based on historical environmental data, historical system operation data, and fault records.

[0010] The present invention also provides a self-cleaning system for an acoustic signal acquisition device, comprising: An acoustic signal acquisition device, including a plurality of microphone channels, each microphone channel being used to monitor acoustic signals; A cleaning subsystem for removing dust and moisture on the surface of the acoustic signal acquisition device; An environmental sensor for collecting environmental data; A controller for controlling the starting time of the cleaning subsystem according to the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device and the environmental data collected by the environmental sensor.

[0011] According to the self-cleaning system for an acoustic signal acquisition device provided by the present invention, the controller includes: A self-cleaning start control module, a self-cleaning prediction module, and a life prediction module; The self-cleaning start control module includes a self-cleaning start control model, which is trained based on the historical audio signal characteristics monitored by each microphone channel, historical environmental data, and historical cleaning records; The self-cleaning prediction module includes a self-cleaning prediction model, which is trained based on historical cleaning records and an environmental change data sequence; The life prediction module includes a life prediction model, which is trained based on historical environmental data, historical system operation data, and fault records.

[0012] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the self-cleaning method for an acoustic signal acquisition device as described in any one of the above.

[0013] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the self-cleaning method for an acoustic signal acquisition device as described in any one of the above.

[0014] The self-cleaning method, device, equipment and storage medium of the sound signal acquisition device provided by the present invention obtain the acoustic signals monitored by each microphone channel in the sound signal acquisition device; extract the time-domain features, frequency-domain features and time-frequency features in the acoustic signals monitored by each microphone channel, input the time-domain features, frequency-domain features and time-frequency features of each microphone channel, the current environmental data and the cleaning record into the self-cleaning start control model, and output the cleaning requirement scores of each microphone channel; start the self-cleaning control for the microphone channels with scores lower than the score threshold according to the cleaning requirement scores of each microphone channel; wherein, the self-cleaning start control model is trained based on the historical audio signal features, historical environmental data and historical cleaning records monitored by each microphone channel, realizing the self-cleaning control of the sound signal acquisition device, avoiding the risks brought by manual maintenance, and accurately predicting the timing of starting self-cleaning, cleaning the specified channels separately, minimizing resource use, and improving the performance and reliability of the sound signal acquisition device. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 It is one of the schematic flowcharts of the self-cleaning method of the sound signal acquisition device provided by the embodiment of the present invention; Figure 2 It is the second schematic flowchart of the self-cleaning method of the sound signal acquisition device provided by the embodiment of the present invention; Figure 3 It is the schematic functional structure diagram of the self-cleaning system of the sound signal acquisition device provided by the embodiment of the present invention; Figure 4 It is the schematic working flowchart of the self-cleaning system of the sound signal acquisition device provided by the embodiment of the present invention; Figure 5 It is the schematic functional structure diagram of the electronic device provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] To make the objectives, technical solutions and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention fall within the protection scope of the present invention.

[0018] Figure 1 The flowchart of the self-cleaning method for the acoustic signal acquisition device provided by the embodiment of the present invention is as follows: Figure 1 As shown, the self-cleaning method for the acoustic signal acquisition device provided by the embodiment of the present invention includes: Step 101, obtaining the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device; Step 102, extracting the time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel, inputting the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records into the self-cleaning start control model, and outputting the cleaning requirement scores for each microphone channel; Step 103, starting the self-cleaning control for the microphone channels with cleaning requirement scores lower than the score threshold according to the cleaning requirement scores of each microphone channel; Among them, the self-cleaning start control model is trained based on the historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel.

[0019] In the embodiment of the present invention, the environmental data includes dust concentration, humidity, temperature, etc.; the historical cleaning records include the cleaning time, cleaning effect, etc. of each microphone channel.

[0020] The traditional microphone array cleaning method is completed by manually replacing the microphone cover regularly. This method is manually operated, with low efficiency. Moreover, since the microphone array has multiple microphone channels and a large area, the dust and moisture contaminated on each microphone channel are different. It is not easy to determine the replacement time of the microphone cover by using the method of uniformly replacing the microphone cover. In addition, since the working environment is underground, manual operation is dangerous.

[0021] The self-cleaning method for the acoustic signal acquisition device provided by the embodiment of the present invention realizes the self-cleaning control of the acoustic signal acquisition device by obtaining the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device; extracting the time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel, inputting the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records into the self-cleaning start control model, and outputting the cleaning requirement scores for each microphone channel; starting the self-cleaning control for the microphone channels with cleaning requirement scores lower than the score threshold according to the cleaning requirement scores of each microphone channel; among them, the self-cleaning start control model is trained based on the historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel, avoiding the danger brought by manual maintenance, and can accurately predict the self-cleaning start timing, clean the specified channel separately, minimize resource use, and improve the performance and reliability of the acoustic signal acquisition device.

[0022] Based on any of the above embodiments, extracting the time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel includes: Extracting the time-domain features from the acoustic signals monitored by each microphone channel includes mean, variance, peak, and kurtosis; Extracting the frequency-domain features from the acoustic signals monitored by each microphone channel includes spectral peak, spectral centroid, and spectral variance; Extracting the time-frequency features from the acoustic signals monitored by each microphone channel includes time-frequency feature wavelet coefficients or energy distribution diagrams.

[0023] In the embodiments of the present invention, inputting the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records into the self-cleaning start control model, and outputting the cleaning requirement scores for each microphone channel includes: Step 201: Fuse the time-domain features, frequency-domain features, and time-frequency features of each microphone channel to generate a multi-dimensional feature vector; Step 202: Combine the current environmental data and historical cleaning records to generate a comprehensive feature matrix; Step 203: Input the multi-dimensional feature vector and the comprehensive feature matrix into the trained self-cleaning start control model, and output the cleaning requirement scores for each microphone channel.

[0024] In the embodiments of the present invention, the self-cleaning start control model analyzes the audio signals of the microphone array to determine which microphone channels need to be cleaned. Its input data includes: audio signal features such as time-domain features: mean, variance, peak, kurtosis, etc.; frequency-domain features: spectral peak, spectral centroid, spectral variance, etc.; time-frequency features: wavelet coefficients, short-time Fourier transform (STFT) energy distribution, etc.

[0025] Environmental data such as dust concentration, humidity, temperature, etc. Historical cleaning records such as the cleaning time and cleaning effect of each microphone channel.

[0026] The output result of the self-cleaning start control model includes the cleaning requirement scores for each microphone channel (such as probability values between 0 and 1). Based on the cleaning requirement scores of each microphone channel, it can be determined which microphone channels need to be cleaned immediately. For example, if the cleaning requirement score of a microphone channel is lower than 0.5, immediate cleaning is initiated.

[0027] In the embodiments of the present invention, the self-cleaning startup control model can be selected from a convolutional neural network (CNN), a support vector machine (SVM), a random forest (RF), etc. When selecting the model, it can be selected according to the characteristics of the model and the actual problems to be solved currently. For example, a convolutional neural network is suitable for extracting the time-frequency characteristics of audio signals. A support vector machine is suitable for small-sample classification problems. A random forest is suitable for classification tasks of multi-feature fusion.

[0028] In the embodiments of the present invention, the training method of the self-cleaning startup control model includes: Step 301: Collect the audio signal data of each microphone channel in the normal state and the polluted state, and label the cleaning requirements; Step 302: Extract the time-domain features, frequency-domain features, and time-frequency domain features from the audio signals, input the time-domain features, frequency-domain features, time-frequency domain features, historical environmental data, and historical cleaning records into the self-cleaning startup control model, and output the predicted cleaning requirement scores of each microphone channel; Step 303: Construct a loss function according to the predicted cleaning requirement scores and the labeled cleaning requirements, adjust the parameters of the self-cleaning startup control model to optimize the loss function until the training end condition is met, and obtain the trained self-cleaning startup control.

[0029] In the embodiments of the present invention, cross-validation is used to evaluate the model performance to ensure the generalization ability. The application scenario of the self-cleaning startup control model provided by the embodiments of the present invention is to monitor the microphone signal in real time and dynamically judge the cleaning requirements. It can reduce unnecessary cleaning operations and reduce energy consumption.

[0030] Based on any of the above embodiments, the self-cleaning method of the sound signal acquisition device further includes: Step 401: Obtain the historical cleaning records and the environmental change data within a period of time; Step 402: Input the historical cleaning records and the environmental change data within a period of time into the self-cleaning prediction model, and output the next self-cleaning startup time; Wherein, the self-cleaning prediction model is trained based on the historical cleaning records and the environmental change data sequence.

[0031] In the embodiments of the present invention, the self-cleaning prediction model can be a long short-term memory network (LSTM), which is suitable for time series data, such as cleaning cycle prediction. The future time when cleaning is required is predicted through the self-cleaning prediction model.

[0032] Based on any of the above embodiments, the self-cleaning method of the sound signal acquisition device further includes: Step 501: Obtain the historical environmental data, system operation data, and fault records; Step 502: Input the historical environmental data, system operation data, and fault records into the life prediction model to obtain the remaining life of the system; Step 503: Output a maintenance plan based on the remaining life of the system; Among them, the life prediction model is trained based on historical environmental data, historical system operation data, and fault records.

[0033] In the embodiment of the present invention, the life prediction model can predict the service life of the self-cleaning system and determine when the self-cleaning system needs to be replaced.

[0034] In the embodiment of the present invention, the input data of the life prediction model includes system operation data such as the working time of the air pump, the number of times the solenoid valve is switched, and the energy consumption; fault records such as the system fault time and the fault type. The output result is, for example, the remaining life prediction of the self-cleaning system (such as the remaining working days).

[0035] In the embodiment of the present invention, the life prediction model can select a regression model such as the Distributed Gradient Boosting Library (XGBoost), Light Gradient Boosting Machine (LightGBM), or a survival analysis model (such as the Cox proportional hazards model), etc.

[0036] The training method of the self-cleaning prediction model or the life prediction model includes collecting historical cleaning records, environmental data, and system operation data. Extract features related to the cleaning cycle and system life, such as the cleaning interval and the environmental deterioration rate. For cleaning cycle prediction, use the LSTM model to learn time series patterns. For life prediction, use a regression model or a survival analysis model. Use historical data to verify the prediction accuracy of the model. The self-cleaning prediction model can optimize the cleaning plan and reduce unnecessary cleaning operations. The life prediction model can give an early warning of the system life and avoid sudden failures.

[0037] In the embodiment of the present invention, by introducing a machine learning model, intelligent cleaning requirement judgment and cleaning cycle prediction can be realized, significantly improving the performance and reliability of the self-cleaning microphone array. The audio signal analysis model is used to judge the cleaning requirement in real time, while the cleaning cycle and life prediction model are used to optimize the maintenance plan and give an early warning of system failures. The combination of these two models enables the system to adapt to complex downhole environments, reduce operation and maintenance costs, and improve work efficiency.

[0038] In some embodiments of the present invention, it also includes using the signal-to-noise ratio and the dust concentration level to determine the cleaning timing, such as Figure 2 As shown, the specific method for determining the cleaning timing includes monitoring the signal-to-noise ratio, frequency response, and other indicators. By comparing the spectral characteristics of the current signal with the historical signal, it is judged whether the signal is attenuated.

[0039] Cleaning decision model: Cleaning requirement = α·SNR + β·dust concentration + γ·humidity Where α, β, and γ are weighting coefficients, which are adjusted according to the actual environment.

[0040] If the performance degradation exceeds the threshold, the cleaning process is initiated using the air flow. The signal quality is re-evaluated after cleaning.

[0041] The self-cleaning method of the acoustic signal acquisition device provided by the embodiments of the present invention provides safety for dangerous environments. In terms of energy management, cleaning is only performed when necessary to minimize energy consumption. Moreover, the system can be compatible with existing power supplies and energy-saving components. A status indicator and a maintenance alarm are provided, enabling remote monitoring and integration with the systems within the mine range; the system has low maintenance costs and can protect personnel safety under certain conditions.

[0042] The self-cleaning system of the acoustic signal acquisition device provided by the present invention is described below. The self-cleaning system of the acoustic signal acquisition device described below can be correspondingly referred to the self-cleaning method of the acoustic signal acquisition device described above.

[0043] Figure 3 is a schematic structural diagram of the self-cleaning system of the acoustic signal acquisition device provided by the embodiments of the present invention, as Figure 3 shown. The self-cleaning system of the acoustic signal acquisition device provided by the embodiments of the present invention includes: An acoustic signal acquisition device, including a plurality of microphone channels, each microphone channel being used to monitor acoustic signals; The acoustic signal acquisition device is, for example, a microphone array module: used to collect acoustic signals, composed of a plurality of microphone units, arranged in a specific geometric shape (such as linear, circular), and using a breathable waterproof material (such as Gore-Tex) to block coal dust and water mist while allowing sound waves to pass through.

[0044] A cleaning subsystem, used to remove dust and moisture on the surface of the acoustic signal acquisition device; In the embodiments of the present invention, the cleaning subsystem: includes an air pump, a gas filtering device, an air storage tank, a barometer, a solenoid valve, and a nozzle, used to purify air and pressurize the clean air for removing coal dust and water mist on the surface of the microphone.

[0045] An environment sensor, used to collect environment data; A controller, used to control the start timing of the cleaning subsystem according to the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device and the environment data collected by the environment sensor.

[0046] In an embodiment of the present invention, the controller is used for signal quality detection, environmental data analysis, cleaning control, and storing historical data and cleaning records. The environmental sensors include a dust sensor and a temperature and humidity sensor, which are used to monitor environmental conditions and transmit sensing information to the controller.

[0047] In an embodiment of the present invention, the controller includes: A self-cleaning start control module, a self-cleaning prediction module, and a lifespan prediction module; The self-cleaning start control module includes a self-cleaning start control model, which is trained based on the historical audio signal characteristics, historical environmental data, and historical cleaning records monitored by each microphone channel; The self-cleaning prediction module includes a self-cleaning prediction model, which is trained based on historical cleaning records and an environmental change data sequence; The lifespan prediction module includes a lifespan prediction model, which is trained based on historical environmental data, historical system operation data, and fault records.

[0048] In some embodiments of the present invention, it further includes: A power supply module: providing power support for the system.

[0049] A communication module: supporting remote data transmission and remote control.

[0050] As Figure 4 shown, the microphone signal and environmental data are collected in real time. The signal quality is detected through a signal processing algorithm. According to the detection result and environmental data, it is decided whether to start the cleaning module. The air pump and solenoid valve are controlled to perform the cleaning operation. After the cleaning is completed, the signal quality is detected again to evaluate the cleaning effect.

[0051] The self-cleaning system of the acoustic signal acquisition device provided by the embodiment of the present invention obtains the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device; extracts the time-domain features, frequency-domain features, and time-frequency features in the acoustic signals monitored by each microphone channel, inputs the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records, into the self-cleaning start control model, and outputs the cleaning requirement score of each microphone channel; based on the cleaning requirement score of each microphone channel, the self-cleaning control is started for the microphone channels with scores lower than the score threshold; wherein, the self-cleaning start control model is trained based on the historical audio signal characteristics, historical environmental data, and historical cleaning records monitored by each microphone channel, realizing the self-cleaning control of the acoustic signal acquisition device, avoiding the danger brought by manual maintenance, and can accurately predict the self-cleaning start timing, perform separate cleaning on the specified channels, minimize resource usage, and improve the performance and reliability of the acoustic signal acquisition device.

[0052] Figure 5 Illustrates a schematic diagram of the physical structure of an electronic device, such as Figure 5 shown. The electronic device may include: a processor 510, a communications interface 520, a memory 530, and a communication bus 540. Among them, the processor 510, the communications interface 520, and the memory 530 complete communication with each other through the communication bus 540. The memory 530 includes computer programs, an operating system, and acquired data. The processor 510 can call the logical instructions in the memory 530 to execute the self-cleaning method of the sound signal acquisition device. The method includes: acquiring the acoustic signals monitored by each microphone channel in the sound signal acquisition device; extracting the time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel, and inputting the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records into the self-cleaning start control model, and outputting the cleaning requirement scores of each microphone channel; starting self-cleaning control for the microphone channels with cleaning requirement scores lower than the score threshold according to the cleaning requirement scores of each microphone channel; where the self-cleaning start control model is trained based on the historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel.

[0053] In addition, when the logical instructions in the above-mentioned memory 530 can be implemented in the form of software functional units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the related technology, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0054] On the other hand, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the self-cleaning method of the sound signal acquisition device provided by the above-mentioned various methods. The method includes: obtaining the acoustic signals monitored by each microphone channel in the sound signal acquisition device; extracting the time-domain features, frequency-domain features, and time-frequency features of the acoustic signals monitored by each microphone channel, and inputting the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records, into the self-cleaning start control model to output the cleaning requirement scores of each microphone channel; starting self-cleaning control for the microphone channels with scores lower than the score threshold according to the cleaning requirement scores of each microphone channel; wherein, the self-cleaning start control model is trained based on the historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel.

[0055] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0056] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the above technical solutions, in essence, or the part that contributes to the related technologies, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A self-cleaning method for a sound signal acquisition device, characterized in that Including: Obtain the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device; Extract the time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel, input the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records, into the self-cleaning start control model, and output the cleaning requirement score for each microphone channel; Based on the cleaning requirement scores of each microphone channel, start the self-cleaning control for the microphone channels with scores lower than the score threshold; Among them, the self-cleaning start control model is trained based on the historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel.

2. The self-cleaning method of the acoustic signal acquisition device according to claim 1, wherein The extracting the time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel includes: The extracting the time-domain features from the acoustic signals monitored by each microphone channel includes mean, variance, peak value, and kurtosis; The extracting the frequency-domain features from the acoustic signals monitored by each microphone channel includes spectral peak value, spectral centroid, and spectral variance; The extracting the time-frequency features from the acoustic signals monitored by each microphone channel includes time-frequency feature wavelet coefficients or energy distribution diagrams.

3. The self-cleaning method of the acoustic signal acquisition device according to claim 2, characterized in that The inputting the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, as well as the current environmental data and cleaning records, into the self-cleaning start control model and outputting the cleaning requirement score for each microphone channel includes: Fuse the time-domain features, frequency-domain features, and time-frequency features of each microphone channel to generate a multi-dimensional feature vector; Combine the current environmental data and historical cleaning records to generate a comprehensive feature matrix; Input the multi-dimensional feature vector and the comprehensive feature matrix into the trained self-cleaning start control model, and output the cleaning requirement score for each microphone channel.

4. The self-cleaning method of the acoustic signal acquisition device according to any one of claims 1 to 3, characterized in that, The training method of the self-cleaning start control model includes: Collect the audio signal data of each microphone channel in the normal state and the contaminated state, and label the cleaning requirements; Extract the time-domain features, frequency-domain features, and time-frequency domain features from the audio signals, input the time-domain features, frequency-domain features, and time-frequency domain features, as well as the historical environmental data and historical cleaning records, into the self-cleaning start control model, and output the predicted cleaning requirement score for each microphone channel; Construct a loss function according to the predicted cleaning requirement score and the labeled cleaning requirement, adjust the parameters of the self-cleaning start control model to optimize the loss function until the training end condition is met, and obtain the trained self-cleaning start control.

5. The self-cleaning method of the acoustic signal acquisition device according to claim 1, characterized in that It also includes: Obtain the historical cleaning records and the environmental change data within a period of time; Input the historical cleaning records and the environmental change data within a period of time into the self-cleaning prediction model, and output the next self-cleaning start time; Among them, the self-cleaning prediction model is trained based on the historical cleaning records and the environmental change data sequence.

6. The self-cleaning method of the acoustic signal acquisition device according to claim 1, characterized in that It also includes: Obtain the historical environmental data, system operation data, and fault records; Input the historical environmental data, system operation data, and fault records into the remaining life prediction model to obtain the remaining life of the system; Output a maintenance plan based on the remaining life of the system; Among them, the life prediction model is trained based on historical environmental data, historical system operation data, and fault records.

7. A self-cleaning system for a sound signal acquisition device, characterized in that, It includes: An acoustic signal acquisition device, including multiple microphone channels, each microphone channel is used to monitor acoustic signals; A cleaning subsystem for removing dust and moisture on the surface of the acoustic signal acquisition device; An environmental sensor for collecting environmental data; A controller for controlling the startup timing of the cleaning subsystem according to the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device and the environmental data collected by the environmental sensor.

8. The self-cleaning system of the acoustic signal acquisition device according to claim 7, characterized in that The controller includes: A self-cleaning startup control module, a self-cleaning prediction module, and a life prediction module; The self-cleaning startup control module includes a self-cleaning startup control model, and the self-cleaning startup control model is trained based on the historical audio signal characteristics, historical environmental data, and historical cleaning records monitored by each microphone channel; The self-cleaning prediction module includes a self-cleaning prediction model, and the self-cleaning prediction model is trained based on historical cleaning records and environmental change data sequences; The life prediction module includes a life prediction model, and the life prediction model is trained based on historical environmental data, historical system operation data, and fault records.

9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the self-cleaning method of the acoustic signal acquisition device according to any one of claims 1 to 6.

10. A non-transitory readable storage medium, on which a computer program is stored, characterized in that, When the computer program is executed by the processor, it implements the self-cleaning method of the acoustic signal acquisition device according to any one of claims 1 to 6.

Images