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

Abstract

The present application relates to the technical field of intelligent mining. Provided are a self-cleaning method and apparatus for an acoustic signal acquisition apparatus, and a device and a storage medium. The method comprises: acquiring an acoustic signal detected by each microphone channel in an acoustic signal acquisition apparatus; extracting a time-domain feature, a frequency-domain feature and a time-frequency feature from the acoustic signal detected by each microphone channel, and inputting into a self-cleaning-initiation control model the time-domain feature, frequency-domain feature and time-frequency feature of each microphone channel, and the current environmental data and cleaning records, so as to output a cleaning requirement score for each microphone channel; and on the basis of the cleaning requirement score for each microphone channel, initiating self-cleaning control on a microphone channel for which the cleaning requirement score is below a score threshold. Self-cleaning control over an acoustic signal acquisition apparatus is realized, dangers arising from manual maintenance are avoided, and the timing for initiating self-cleaning can be accurately predicted, and specified channels can be cleaned separately, thereby using less resources as much as possible, and improving the performance and reliability of the acoustic signal acquisition apparatus.

Description

Self-cleaning method, device and equipment of sound signal collection device and storage medium

[0001] The present application claims priority to the Chinese patent application No. 2025102237999, filed on February 27, 2025, and entitled "Self-cleaning method, device and equipment of sound signal collection device and storage medium", the whole content of which is incorporated herein by reference. TECHNICAL FIELD

[0002] The present application relates to the technical field of intelligent mining, in particular to a self-cleaning method, device and equipment of sound signal collection device and storage medium. BACKGROUND

[0003] In underground mining operations, sound signal collection 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 prone to decline, and therefore, the microphone array needs to be cleaned. The traditional method of cleaning the microphone array is to replace the microphone cover by manual operation periodically. This method is inefficient and dangerous because the underground working environment is dangerous for manual operation. In addition, since the microphone array has multiple microphone channels and a large area, the dust and moisture adhered to each microphone channel are different. It is not easy to determine the replacement time of the microphone cover by using the uniform replacement method. SUMMARY

[0004] The present application provides a self-cleaning method, device and equipment of sound signal collection device and storage medium to solve the problem that the traditional sound signal collection device is not easy to determine the replacement time of the microphone cover and is dangerous.

[0005] The present application provides a self-cleaning method of sound signal collection device, comprising:

[0006] Obtaining the acoustic signal monitored by each microphone channel in the sound signal collection device;

[0007] Extracting the time domain feature, frequency domain feature and time-frequency feature in the acoustic signal monitored by each microphone channel, inputting the time domain feature, frequency domain feature and time-frequency feature of each microphone channel, current environmental data and cleaning record into a self-cleaning start control model, and outputting the cleaning demand score of each microphone channel;

[0008] Starting self-cleaning control for the microphone channel below the score threshold according to the cleaning demand score of each microphone channel.

[0009] The self-cleaning start control model is trained based on the historical audio signal features, historical environmental data and historical cleaning record monitored by each microphone channel.

[0010] According to the self-cleaning method for an acoustic signal acquisition device provided by the present invention, the extraction of time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel includes:

[0011] Temporal features, including mean, variance, peak value, and kurtosis, are extracted from the acoustic signals monitored by each microphone channel.

[0012] Frequency domain features, including spectral peaks, spectral centroids, and spectral variance, are extracted from the acoustic signals monitored by each microphone channel.

[0013] Extract the time-frequency features from the acoustic signals monitored by each microphone channel, including time-frequency feature wavelet coefficients or energy distribution maps.

[0014] According to the self-cleaning method for an acoustic signal acquisition device provided by the present invention, the step of inputting the time-domain characteristics, frequency-domain characteristics, and time-frequency characteristics of each microphone channel, as well as current environmental data and cleaning records, into a self-cleaning initiation control model, and outputting a cleaning requirement score for each microphone channel, includes:

[0015] The temporal, frequency, and time-frequency features of each microphone channel are fused to generate a multidimensional feature vector.

[0016] A comprehensive feature matrix is ​​generated by combining current environmental data and historical cleaning records;

[0017] The multidimensional feature vector and the comprehensive feature matrix are input into the trained self-cleaning initiation control model, which outputs a cleaning requirement score for each microphone channel.

[0018] According to the self-cleaning method for an acoustic signal acquisition device provided by the present invention, the training method for the self-cleaning start-up control model includes:

[0019] Collect audio signal data from each microphone channel under normal and contaminated conditions, and label cleaning requirements;

[0020] Extract time-domain features, frequency-domain features, and time-frequency-domain features from the audio signal. Input the time-domain features, frequency-domain features, and time-frequency-domain features, along with historical environmental data and historical cleaning records, into the self-cleaning initiation control model and output a predicted cleaning requirement score for each microphone channel.

[0021] A loss function is constructed based on the predicted cleaning demand score and the labeled cleaning demand. The parameters of the self-cleaning initiation control model are adjusted to optimize the loss function until the training termination condition is met, thus obtaining a well-trained self-cleaning initiation control.

[0022] The self-cleaning method for an acoustic signal acquisition device provided by the present invention further includes:

[0023] Acquire historical cleaning records and environmental change data over a period of time;

[0024] Input the historical cleaning records and environmental change data over a period of time into the self-cleaning prediction model, and output the next self-cleaning start time;

[0025] The self-cleaning prediction model is trained based on historical cleaning records and environmental change data sequences.

[0026] The self-cleaning method for an acoustic signal acquisition device provided by the present invention further includes:

[0027] Acquire historical environmental data, system operation data, and fault records;

[0028] Input the historical environmental data, system operation data, and fault records into the lifespan prediction model to obtain the remaining lifespan of the system;

[0029] Output a maintenance plan based on the remaining lifespan of the system;

[0030] The life prediction model is trained based on historical environmental data, historical system operation data, and fault records.

[0031] The present invention also provides a self-cleaning system for an acoustic signal acquisition device, comprising:

[0032] The acoustic signal acquisition device includes multiple microphone channels, each microphone channel being used to monitor acoustic signals;

[0033] The cleaning subsystem is used to remove dust and moisture from the surface of the acoustic signal acquisition device;

[0034] Environmental sensors are used to collect environmental data;

[0035] The controller is used to control the start-up timing of the cleaning subsystem based on the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device and the environmental data acquired by the environmental sensors.

[0036] According to the self-cleaning system for acoustic signal acquisition devices provided by the present invention, the controller includes:

[0037] Self-cleaning start-up control module, self-cleaning prediction module, and lifespan prediction module;

[0038] The self-cleaning start control module includes a self-cleaning start control model, which is trained based on historical audio signal characteristics, historical environmental data, and historical cleaning records monitored by each microphone channel.

[0039] The self-cleaning prediction module includes a self-cleaning prediction model, which is trained based on historical cleaning records and environmental change data sequences.

[0040] The lifespan prediction module includes a lifespan prediction model, which is trained based on historical environmental data, historical system operation data, and fault records.

[0041] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the self-cleaning method for the acoustic signal acquisition device as described in any of the preceding claims.

[0042] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the self-cleaning method of the acoustic signal acquisition device described in any of the preceding claims.

[0043] The self-cleaning method, apparatus, device, and storage medium for an acoustic signal acquisition device provided by this invention acquires 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 from the acoustic signals monitored by each microphone channel; inputs the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, along with current environmental data and cleaning records, into a self-cleaning initiation control model; and outputs a cleaning requirement score for each microphone channel. Based on the cleaning requirement score of each microphone channel, self-cleaning control is initiated for microphone channels with scores below a threshold. The self-cleaning initiation control model is trained based on historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel, realizing self-cleaning control of the acoustic signal acquisition device, avoiding the dangers of manual maintenance, and accurately predicting the timing of self-cleaning initiation. It allows for individual cleaning of designated channels, minimizing resource usage and improving the performance and reliability of the acoustic signal acquisition device.

[0044] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 is one of the flowcharts of the self-cleaning method for the acoustic signal acquisition device provided in an embodiment of the present invention;

[0047] Figure 2 is a second schematic flowchart of the self-cleaning method for the acoustic signal acquisition device provided in an embodiment of the present invention;

[0048] Figure 3 is a schematic diagram of the functional structure of the self-cleaning system of the acoustic signal acquisition device provided in an embodiment of the present invention;

[0049] Figure 4 is a schematic diagram of the working process of the self-cleaning system of the acoustic signal acquisition device provided in an embodiment of the present invention;

[0050] Figure 5 is a functional structure diagram of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0051] The embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate this application, but should not be used to limit the scope of this application.

[0052] In the description of the embodiments of this application, it should be noted that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of this application. In addition, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0053] In the description of the embodiments of this application, it should be noted that, unless otherwise explicitly specified and limited, the terms "connected" and "linked" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in the embodiments of this application based on the specific circumstances.

[0054] In the embodiments of this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0055] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the embodiments of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0056] Figure 1 is a flowchart of the self-cleaning method for an acoustic signal acquisition device provided in an embodiment of the present invention. As shown in Figure 1, the self-cleaning method for an acoustic signal acquisition device provided in an embodiment of the present invention includes:

[0057] Step 101: Acquire the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device;

[0058] Step 102: 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, along with the current environmental data and cleaning records, into the self-cleaning start-up control model and output the cleaning requirement score for each microphone channel.

[0059] Step 103: Based on the cleaning requirement score of each microphone channel, initiate self-cleaning control for microphone channels that are below the score threshold;

[0060] The self-cleaning start-up control model is trained based on historical audio signal characteristics, historical environmental data, and historical cleaning records monitored by each microphone channel.

[0061] In this embodiment of the invention, environmental data includes dust concentration, humidity, temperature, etc.; historical cleaning records include cleaning time and cleaning effect of each microphone channel, etc.

[0062] Traditional microphone array cleaning methods involve manually replacing microphone covers periodically. This method is inefficient due to its manual operation. Furthermore, since microphone arrays have multiple microphone channels and a large area, the amount of dust and moisture accumulated in each microphone channel varies. It is difficult to determine the replacement time of microphone covers by replacing them uniformly. In addition, manual operation is dangerous because the work environment is underground.

[0063] The self-cleaning method for an acoustic signal acquisition device provided in this invention involves acquiring 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, along with current environmental data and cleaning records, into a self-cleaning initiation control model; and outputting a cleaning requirement score for each microphone channel. Based on the cleaning requirement score of each microphone channel, self-cleaning control is initiated for microphone channels with scores below a threshold. The self-cleaning initiation control model is trained based on historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel. This achieves self-cleaning control of the acoustic signal acquisition device, avoids the dangers of manual maintenance, and can accurately predict the timing of self-cleaning initiation, performing individual cleaning on designated channels, minimizing resource usage, and improving the performance and reliability of the acoustic signal acquisition device.

[0064] Based on any of the above embodiments, the extraction of time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel includes:

[0065] Temporal features, including mean, variance, peak value, and kurtosis, are extracted from the acoustic signals monitored by each microphone channel.

[0066] Frequency domain features, including spectral peaks, spectral centroids, and spectral variance, are extracted from the acoustic signals monitored by each microphone channel.

[0067] Extract the time-frequency features from the acoustic signals monitored by each microphone channel, including time-frequency feature wavelet coefficients or energy distribution maps.

[0068] In this embodiment of the invention, the step of inputting the time-domain characteristics, frequency-domain characteristics, and time-frequency characteristics of each microphone channel, along with current environmental data and cleaning records, into the self-cleaning initiation control model, and outputting a cleaning requirement score for each microphone channel, includes:

[0069] 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;

[0070] Step 202: Combine current environmental data and historical cleaning records to generate a comprehensive feature matrix;

[0071] Step 203: Input the multidimensional feature vector and the comprehensive feature matrix into the trained self-cleaning start-up control model, and output the cleaning requirement score for each microphone channel.

[0072] In this embodiment of the invention, the self-cleaning start-up control model determines which microphone channels need cleaning by analyzing the audio signals of the microphone array. Its input data includes: audio signal characteristics such as time-domain characteristics (mean, variance, peak value, kurtosis, etc.); frequency-domain characteristics (spectral peak value, spectral centroid, spectral variance, etc.); and time-frequency characteristics (wavelet coefficients, short-time Fourier transform (STFT) energy distribution, etc.).

[0073] Environmental data such as dust concentration, humidity, and temperature. Historical cleaning records such as cleaning time and cleaning effect for each microphone channel.

[0074] The output of the self-cleaning start control model includes a cleaning requirement score for each microphone channel (such as a probability value between 0 and 1). Based on the cleaning requirement score 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 less than 5, immediate cleaning can be initiated.

[0075] In this embodiment of the invention, the self-cleaning startup control model can be selected from convolutional neural networks (CNN), support vector machines (SVM), and random forests (RF), etc. The model selection can be based on its characteristics and the actual problem to be solved. For example, convolutional neural networks are suitable for extracting time-frequency features of audio signals. Support vector machines are suitable for small-sample classification problems. Random forests are suitable for classification tasks involving multi-feature fusion.

[0076] In this embodiment of the invention, the training method for the self-cleaning start-up control model includes:

[0077] Step 301: Collect audio signal data for each microphone channel under normal and contaminated conditions, and label the cleaning requirements;

[0078] Step 302: Extract time-domain features, frequency-domain features, and time-frequency-domain features from the audio signal. Input the time-domain features, frequency-domain features, and time-frequency-domain features, along with historical environmental data and historical cleaning records, into the self-cleaning start-up control model and output the predicted cleaning requirement score for each microphone channel.

[0079] Step 303: Construct a loss function based on the predicted cleaning demand score and the labeled cleaning demand, adjust the parameters of the self-cleaning start control model to optimize the loss function until the training termination condition is met, and obtain the trained self-cleaning start control.

[0080] In this embodiment of the invention, cross-validation is used to evaluate model performance and ensure generalization ability. The self-cleaning startup control model provided in this embodiment of the invention is used in the application scenario of real-time monitoring of microphone signals to dynamically determine cleaning needs. It can reduce unnecessary cleaning operations and lower energy consumption.

[0081] Based on any of the above embodiments, the self-cleaning method for the acoustic signal acquisition device further includes:

[0082] Step 401: Obtain historical cleaning records and environmental change data over a period of time;

[0083] Step 402: Input the historical cleaning records and environmental change data over a period of time into the self-cleaning prediction model, and output the next self-cleaning start time;

[0084] The self-cleaning prediction model is trained based on historical cleaning records and environmental change data sequences.

[0085] In this embodiment of the invention, the self-cleaning prediction model can be a Long Short-Term Memory (LSTM) network, which is suitable for time series data, such as cleaning cycle prediction. The self-cleaning prediction model predicts when cleaning is needed in the future.

[0086] Based on any of the above embodiments, the self-cleaning method for the acoustic signal acquisition device further includes:

[0087] Step 501: Obtain historical environmental data, system operation data, and fault records;

[0088] Step 502: Input the historical environmental data, system operation data, and fault records into the lifespan prediction model to obtain the remaining lifespan of the system;

[0089] Step 503: Output a maintenance plan based on the remaining lifespan of the system;

[0090] The life prediction model is trained based on historical environmental data, historical system operation data, and fault records.

[0091] In this embodiment of the invention, the life prediction model can predict the lifespan of the self-cleaning system and determine when the self-cleaning system needs to be replaced.

[0092] In this embodiment of the invention, the input data for the life prediction model includes system operating data such as air pump operating time, solenoid valve switching frequency, energy consumption, etc.; and fault records such as system fault time and fault type. Its output result is, for example, a prediction of the remaining lifespan of the self-cleaning system (e.g., remaining working days).

[0093] In this embodiment of the invention, the life prediction model can be a regression model such as the distributed gradient boosting library (XGBoost), the lightweight gradient boosting machine (LightGBM), or a survival analysis model (such as the Cox proportional hazards model).

[0094] Training methods for self-cleaning prediction models or lifespan prediction models include collecting historical cleaning records, environmental data, and system operation data. Features related to cleaning cycles and system lifespan, such as cleaning intervals and environmental degradation rates, are extracted. For cleaning cycle prediction, an LSTM model is used to learn time series patterns. For lifespan prediction, a regression model or survival analysis model is used. Historical data is used to validate the model's predictive accuracy. Self-cleaning prediction models can optimize cleaning plans and reduce unnecessary cleaning operations. Lifespan prediction models can provide early warnings of system lifespan, preventing sudden failures.

[0095] In this embodiment of the invention, the introduction of a machine learning model enables intelligent judgment of cleaning needs and prediction of cleaning cycles, significantly improving the performance and reliability of the self-cleaning microphone array. An audio signal analysis model is used to determine cleaning needs in real time, while a cleaning cycle and lifespan prediction model is used to optimize maintenance plans and provide early warnings of system failures. The combination of these two models allows the system to adapt to complex downhole environments, reduce operation and maintenance costs, and improve work efficiency.

[0096] In some embodiments of the present invention, the timing of cleaning is determined using signal-to-noise ratio (SNR) and dust concentration levels, as shown in Figure 2. The specific method for determining the timing of cleaning includes monitoring SNR, frequency response, and other indicators. Signal attenuation is determined by comparing the spectral characteristics of the current signal with those of historical signals.

[0097] Cleaning decision model: Cleaning demand = α·SNR + β·dust concentration + γ·humidity

[0098] Among them, α, β, and γ are weighting coefficients, which are adjusted according to the actual environment.

[0099] If performance degradation exceeds a threshold, a cleaning process is initiated using airflow. Signal quality is then reassessed after cleaning.

[0100] The self-cleaning method for acoustic signal acquisition devices provided by this invention offers safety in hazardous environments. In terms of energy management, cleaning is performed only when necessary to minimize energy consumption. Furthermore, the system is compatible with existing power supplies and energy-saving components. It provides status indicators and maintenance alarms, enables remote monitoring, and can be integrated with systems across the mine; the system has low maintenance costs and can protect personnel safety under certain conditions.

[0101] The self-cleaning system for the acoustic signal acquisition device provided by the present invention is described below. The self-cleaning system for the acoustic signal acquisition device described below and the self-cleaning method for the acoustic signal acquisition device described above can be referred to in correspondence with each other.

[0102] Figure 3 is a structural schematic diagram of the self-cleaning system of the acoustic signal acquisition device provided in an embodiment of the present invention. As shown in Figure 3, the self-cleaning system of the acoustic signal acquisition device provided in an embodiment of the present invention includes:

[0103] The acoustic signal acquisition device includes multiple microphone channels, each microphone channel being used to monitor acoustic signals;

[0104] Sound signal acquisition devices, for example, are microphone array modules: used to acquire sound signals, consisting of multiple microphone units arranged in a specific geometry (such as linear or circular), and using breathable and waterproof materials (such as Gore-Tex) to block coal dust and water mist while allowing sound waves to pass through.

[0105] The cleaning subsystem is used to remove dust and moisture from the surface of the acoustic signal acquisition device;

[0106] In this embodiment of the invention, the cleaning subsystem includes an air pump, a gas filter, an air tank, a barometer, a solenoid valve, and a nozzle, for purifying the air and pressurizing the clean air to remove coal dust and water mist from the microphone surface.

[0107] Environmental sensors are used to collect environmental data;

[0108] The controller is used to control the start-up timing of the cleaning subsystem based on the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device and the environmental data acquired by the environmental sensors.

[0109] In this embodiment of the invention, the controller is used for signal quality inspection, environmental data analysis, cleaning control, and storing historical data and cleaning records. Environmental sensors, including dust sensors and temperature and humidity sensors, are used to monitor environmental conditions and transmit the sensing information to the controller.

[0110] In this embodiment of the invention, the controller includes:

[0111] Self-cleaning start-up control module, self-cleaning prediction module, and lifespan prediction module;

[0112] The self-cleaning start control module includes a self-cleaning start control model, which is trained based on historical audio signal characteristics, historical environmental data, and historical cleaning records monitored by each microphone channel.

[0113] The self-cleaning prediction module includes a self-cleaning prediction model, which is trained based on historical cleaning records and environmental change data sequences.

[0114] The lifespan prediction module includes a lifespan prediction model, which is trained based on historical environmental data, historical system operation data, and fault records.

[0115] In some embodiments of the present invention, it further includes:

[0116] Power module: Provides power support for the system.

[0117] Communication module: Supports remote data transmission and remote control.

[0118] As shown in Figure 4, microphone signals and environmental data are acquired in real time. Signal quality is detected using a signal processing algorithm. Based on the detection results and environmental data, a decision is made whether to activate the cleaning module. The air pump and solenoid valve are controlled to perform the cleaning operation. After cleaning is completed, the signal quality is detected again to evaluate the cleaning effect.

[0119] The self-cleaning system for an acoustic signal acquisition device provided in this invention acquires 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 from the acoustic signals monitored by each microphone channel; inputs the time-domain features, frequency-domain features, and time-frequency features of each microphone channel, along with current environmental data and cleaning records, into a self-cleaning start-up control model; and outputs a cleaning requirement score for each microphone channel. Based on the cleaning requirement score of each microphone channel, self-cleaning control is initiated for microphone channels with scores below a threshold. The self-cleaning start-up control model is trained based on historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel, thus realizing self-cleaning control of the acoustic signal acquisition device, avoiding the dangers of manual maintenance, and accurately predicting the timing of self-cleaning initiation. It allows for individual cleaning of designated channels, minimizing resource usage and improving the performance and reliability of the acoustic signal acquisition device.

[0120] Figure 5 illustrates a schematic diagram of the physical structure of an electronic device. As shown in Figure 5, the electronic device may include: a processor 510, a communication interface 520, a memory 530, and a communication bus 540. The processor 510, the communication interface 520, and the memory 530 communicate with each other through the communication bus 540. The memory 530 includes a computer program, an operating system, and acquired data. The processor 510 can call the logical instructions in the memory 530 to execute a self-cleaning method for the acoustic signal acquisition device. The method includes: acquiring 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, along with current environmental data and cleaning records, into a self-cleaning initiation control model; and outputting a cleaning requirement score for each microphone channel; and initiating self-cleaning control for microphone channels with cleaning requirement scores below a threshold based on the cleaning requirement scores of each microphone channel. The self-cleaning initiation control model is trained based on historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel.

[0121] Furthermore, the logical instructions in the aforementioned memory 530 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to related technologies, or a portion of the 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 to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0122] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the self-cleaning method for an acoustic signal acquisition device provided by the methods described above. The method includes: acquiring acoustic signals monitored by each microphone channel in the acoustic signal acquisition device; extracting 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, along with current environmental data and cleaning records, into a self-cleaning initiation control model; and outputting a cleaning requirement score for each microphone channel; initiating self-cleaning control for microphone channels with cleaning requirement scores below a threshold based on the cleaning requirement scores of each microphone channel; wherein the self-cleaning initiation control model is trained based on historical audio signal features, historical environmental data, and historical cleaning records monitored by each microphone channel.

[0123] The above are merely preferred embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A self-cleaning method for an acoustic signal acquisition device, characterized in that, include: Acquire the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device; Extract the time-domain, frequency-domain, and time-frequency features from the acoustic signals monitored by each microphone channel. Input the time-domain, frequency-domain, and time-frequency features of each microphone channel, along with the current environmental data and cleaning records, into the self-cleaning initiation control model and output a cleaning requirement score for each microphone channel. Self-cleaning control is initiated for microphone channels that are below the cleaning requirement threshold based on the cleaning requirement score for each microphone channel. The self-cleaning start-up control model is trained based on historical audio signal characteristics, historical environmental data, and historical cleaning records monitored by each microphone channel.

2. The self-cleaning method for the acoustic signal acquisition device according to claim 1, characterized in that, The extraction of time-domain features, frequency-domain features, and time-frequency features from the acoustic signals monitored by each microphone channel includes: Temporal features, including mean, variance, peak value, and kurtosis, are extracted from the acoustic signals monitored by each microphone channel. Frequency domain features, including spectral peaks, spectral centroids, and spectral variance, are extracted from the acoustic signals monitored by each microphone channel. Extract the time-frequency features from the acoustic signals monitored by each microphone channel, including time-frequency feature wavelet coefficients or energy distribution maps.

3. The self-cleaning method for the acoustic signal acquisition device according to claim 2, characterized in that, The self-cleaning initiation control model inputs the time-domain characteristics, frequency-domain characteristics, and time-frequency characteristics of each microphone channel, along with current environmental data and cleaning records, and outputs a cleaning requirement score for each microphone channel, including: The temporal, frequency, and time-frequency features of each microphone channel are fused to generate a multidimensional feature vector. A comprehensive feature matrix is ​​generated by combining current environmental data and historical cleaning records; The multidimensional feature vector and the comprehensive feature matrix are input into the trained self-cleaning initiation control model, which outputs a cleaning requirement score for each microphone channel.

4. The self-cleaning method for the acoustic signal acquisition device according to any one of claims 1 to 3, characterized in that, The training method for the self-cleaning start-up control model includes: Collect audio signal data from each microphone channel under normal and contaminated conditions, and label cleaning requirements; Extract time-domain features, frequency-domain features, and time-frequency-domain features from the audio signal. Input the time-domain features, frequency-domain features, and time-frequency-domain features, along with historical environmental data and historical cleaning records, into the self-cleaning initiation control model and output a predicted cleaning requirement score for each microphone channel. A loss function is constructed based on the predicted cleaning demand score and the labeled cleaning demand. The parameters of the self-cleaning initiation control model are adjusted to optimize the loss function until the training termination condition is met, thus obtaining a well-trained self-cleaning initiation control.

5. The self-cleaning method for the acoustic signal acquisition device according to claim 1, characterized in that, Also includes: Acquire historical cleaning records and environmental change data over a period of time; Input the historical cleaning records and environmental change data over a period of time into the self-cleaning prediction model, and output the next self-cleaning start time; The self-cleaning prediction model is trained based on historical cleaning records and environmental change data sequences.

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

7. A self-cleaning system for an acoustic signal acquisition device, characterized in that, include: The acoustic signal acquisition device includes multiple microphone channels, each microphone channel being used to monitor acoustic signals; The cleaning subsystem is used to remove dust and moisture from the surface of the acoustic signal acquisition device; Environmental sensors are used to collect environmental data; The controller is used to control the start-up timing of the cleaning subsystem based on the acoustic signals monitored by each microphone channel in the acoustic signal acquisition device and the environmental data acquired by the environmental sensors.

8. The self-cleaning system of the acoustic signal acquisition device according to claim 7, characterized in that, The controller includes: Self-cleaning start-up control module, self-cleaning prediction module, and lifespan prediction module; The self-cleaning start control module includes a self-cleaning start control model, which is trained based on 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 environmental change data sequences. The lifespan prediction module includes a lifespan prediction model, which 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 in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the self-cleaning method for the acoustic signal acquisition device as described in any one of claims 1 to 6.

10. A non-transitory readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the self-cleaning method for the acoustic signal acquisition device as described in any one of claims 1 to 6.

Images