Diagnostic system and method for audio equipment

Abstract

The embodiment of the invention provides a diagnosis system and method for audio equipment, and the system comprises a sensor module which is configured to collect multi-dimensional sensing data of the audio equipment; the audio acquisition module is configured to acquire an output signal of audio equipment, perform feature analysis on the output signal and determine a signal feature of the output signal; the data transmission module is configured to perform anomaly analysis on the multi-dimensional sensing data, determine abnormal sensing data and transmit the signal characteristics and the abnormal sensing data to the intelligent diagnosis module; and the intelligent diagnosis module is configured to perform abnormality diagnosis on the audio equipment based on the abnormal sensing data and the signal characteristics.

Description

Technical Field

[0001] This specification relates to the field of audio device diagnostic technology, and in particular to a diagnostic system and method for audio devices. Background Technology

[0002] With the increasing popularity of smart audio devices (such as speakers, headphones, amplifiers, and smart speakers), the stability of device operation and sound quality have become important indicators of user experience and product competitiveness.

[0003] Most existing audio equipment diagnostic technologies rely on a single monitoring dimension, such as judging the equipment's operating status solely based on temperature, current, or vibration signals, or identifying faults by analyzing the distortion and noise characteristics of the audio output signal. These methods suffer from limited diagnostic coverage, sensitivity to environmental noise, and an inability to adapt to changes in the operating conditions of audio equipment, easily leading to misdiagnosis or missed diagnosis, thus reducing the accuracy and reliability of the diagnosis.

[0004] Therefore, there is an urgent need for a diagnostic system and method for audio equipment that can integrate audio signals with multi-dimensional sensor data to achieve a comprehensive, accurate, and dynamic health assessment of the operating status of audio equipment. Summary of the Invention

[0005] This specification provides one or more embodiments of a diagnostic system for an audio device. The diagnostic system includes: a sensor module configured to acquire multidimensional sensing data from the audio device; an audio acquisition module configured to acquire the output signal of the audio device and perform feature analysis on the output signal to determine the signal characteristics of the output signal; a data transmission module configured to perform anomaly analysis on the multidimensional sensing data to determine abnormal sensing data and transmit the signal characteristics and the abnormal sensing data to an intelligent diagnostic module; and the intelligent diagnostic module configured to perform anomaly diagnosis on the audio device based on the abnormal sensing data and the signal characteristics.

[0006] This specification provides one or more embodiments of a diagnostic method for an audio device, the diagnostic method comprising: acquiring multidimensional sensor data of the audio device; acquiring the output signal of the audio device and performing feature analysis on the output signal to determine the signal characteristics of the output signal; performing anomaly analysis on the multidimensional sensor data to determine abnormal sensor data; and performing anomaly diagnosis on the audio device based on the abnormal sensor data and the signal characteristics. Attached Figure Description

[0007] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein: Figure 1 This is a schematic diagram illustrating an application scenario of a diagnostic system for audio devices according to some embodiments of this specification; Figure 2 This is an exemplary system block diagram of a diagnostic system for an audio device according to some embodiments of this specification; Figure 3 This is an exemplary flowchart of a diagnostic method for an audio device according to some embodiments of this specification; Figure 4 This is a schematic diagram of an analytical model shown according to some embodiments of this specification; Figure 5 This is an exemplary flowchart illustrating the determination of graded repair parameters according to some embodiments of this specification. Detailed Implementation

[0008] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0009] It should be understood that the terms “system,” “device,” “unit,” and / or “module” used herein are one way to distinguish different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they may be replaced by other expressions.

[0010] As indicated in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0011] Flowcharts are used in this specification to illustrate the operations performed by the system according to embodiments of this specification. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0012] Figure 1 This is a schematic diagram illustrating an application scenario of a diagnostic system for an audio device, based on some embodiments of this specification.

[0013] In some embodiments, the audio device diagnostic system can achieve real-time monitoring and intelligent diagnosis of the audio device's operating status. Its applications are wide-ranging, including cinemas, studios, smart home environments, and commercial displays. The audio device diagnostic system and intelligent diagnostic technology can integrate audio signals with multi-dimensional sensor data to achieve a comprehensive, accurate, and dynamic health assessment of the audio device's operating status, thereby ensuring stable operation and long-term reliability of the device.

[0014] In some embodiments, such as Figure 1 As shown, the application scenario of the diagnostic system for audio equipment (hereinafter referred to as application scenario 100) may include audio equipment 110, user terminal 120, network 130, processor 140 and database 150.

[0015] The audio device 110 can perform functions such as acquiring, processing, and / or outputting sound and other data. In some embodiments, the audio device 110 may include devices such as speakers, digital media players, and audio equipment. In some embodiments, the audio device 110 may be a combination of one or more devices with music playback capabilities.

[0016] User terminal 120 can be used to interact with users. Users can be one or more users, such as users directly using the diagnostic system or other related users. In some embodiments, user terminal 120 may include a mobile phone 121, a tablet 122, and a computer 123, etc.

[0017] Network 130 may include any suitable network capable of facilitating information and / or data exchange. In some embodiments, at least one component of application scenario 100 (e.g., audio device 110, user terminal 120, processor 140, database 150, etc.) may exchange information and / or data with at least one other component in application scenario 100 via network 130. For example, processor 140 may obtain multidimensional sensor data, audio device output signals, etc., from database 150 via network 130.

[0018] In some embodiments, network 130 can be any one or more of wired or wireless networks. For example, network 130 may include cable networks, fiber optic networks, telecommunications networks, cable connections, or any combination thereof. Network connections between components may employ one or more of the above methods. In some embodiments, the network may be a point-to-point, shared, centralized, or other topologies, or a combination of multiple topologies. In some embodiments, network 130 may include one or more network access points.

[0019] Processor 140 can process data and / or information obtained from other devices or system components. Processor 140 can execute program instructions based on this data, information, and / or processing results to perform one or more functions described in this specification. In some embodiments, processor 140 may include one or more sub-processing devices (e.g., a single-core processing device or a multi-core multi-chip processing device). By way of example only, processor 140 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), a microprocessor, or any combination thereof.

[0020] Database 150 may store data, instructions, and / or any other information related to the diagnostic system of the audio device. In some embodiments, database 150 may store data and / or information acquired by the audio device 110, user terminal 120, and processor 140 (e.g., the output signal of the audio device, signal characteristics of the output signal, multidimensional sensor data, abnormal sensor data, etc.).

[0021] In some embodiments, database 150 may store data and / or instructions used by processor 140 to execute or use in order to perform the exemplary methods described herein. For example, database 150 may store diagnostic results of processor 140 performing anomaly diagnosis on an audio device based on abnormal sensing data and signal characteristics.

[0022] In some embodiments, database 150 may include one or more storage units, each of which may be a separate device or part of another device. In some embodiments, database 150 may be implemented on a cloud platform. In some embodiments, database 150 may be part of user terminal 120, processor 140, and / or audio device 110.

[0023] Figure 2 This is an exemplary system block diagram of a diagnostic system for an audio device according to some embodiments of this specification.

[0024] In some embodiments, such as Figure 2 As shown, the audio device diagnostic system 200 may include a sensor module 210, an audio acquisition module 220, a data transmission module 230, and an intelligent diagnostic module 240. A processor 140 may be integrated into one or more of these modules; conversely, one or more of these modules may be integrated into the processor 140.

[0025] In some embodiments, the diagnostic system 200 can be implemented in various forms. For example, the diagnostic system 200 can be a standalone hardware entity connected to the audio device 110 via an interface. Alternatively, the diagnostic system 200 can be a software system, integrated into the processor of the audio device 110 as software or firmware.

[0026] Sensor module 210 refers to a module that collects multi-dimensional sensing data during the operation of the audio device. Multi-dimensional sensing data may include temperature data, current data, voltage data, vibration data, and humidity data, etc. In some embodiments, the multi-dimensional sensing data can reflect the operating status and environmental characteristics of the audio device. In some embodiments, sensor module 210 includes various monitoring and sensing devices, such as temperature sensors, current sensors, voltage sensors, vibration sensors, triaxial sensors, and humidity sensors.

[0027] In some embodiments, the sensor module 210 is communicatively connected to the data transmission module 230.

[0028] In some embodiments, the sensor module 210 is configured to acquire multidimensional sensing data from the audio device.

[0029] The audio acquisition module 220 refers to a module that acquires the output signal of an audio device and analyzes the output signal. In some embodiments, the audio acquisition module 220 may include a microphone array, a digital signal processor, an audio analysis circuit, and an analog-to-digital conversion (ADC) module, etc.

[0030] In some embodiments, the audio acquisition module 220 is communicatively connected to the data transmission module 230.

[0031] In some embodiments, the audio acquisition module 220 is configured to acquire the output signal of the audio device and perform feature analysis on the output signal to determine the signal characteristics of the output signal.

[0032] The data transmission module 230 refers to the module responsible for data analysis and transmission. In some embodiments, the data transmission module may include a Wi-Fi module, a Bluetooth module, a controller area network bus, etc.

[0033] In some embodiments, the data transmission module 230 is communicatively connected to the sensor module 210, the audio acquisition module 220, and the intelligent diagnostic module 240, respectively.

[0034] In some embodiments, the data transmission module is configured to perform anomaly analysis on multidimensional sensing data, identify abnormal sensing data, and transmit signal characteristics and abnormal sensing data to the intelligent diagnostic module.

[0035] The intelligent diagnostic module 240 refers to a module used for diagnosing audio devices. In some embodiments, the intelligent diagnostic module includes a server, a processor, a graphics processing unit (GPU), etc.

[0036] In some embodiments, the intelligent diagnostic module is configured to perform anomaly diagnosis on the audio device based on abnormal sensing data and signal characteristics.

[0037] In some embodiments, the intelligent diagnostic module is further configured to: determine the anomaly probability distribution based on abnormal sensing data and signal characteristics through an analysis model; the analysis model is a machine learning model; and perform anomaly diagnosis on the audio device based on the anomaly probability distribution.

[0038] In some embodiments, the intelligent diagnostic module is further configured to: determine sensing spatial features and signal spatial features through a feature extraction layer based on abnormal sensing data and signal features; and determine the abnormal probability distribution through an anomaly analysis layer based on the sensing spatial features and signal spatial features.

[0039] In some embodiments, the intelligent diagnostic module is further configured to determine reference output characteristics based on playback parameters and input signal characteristics.

[0040] In some embodiments, the intelligent diagnostic module is further configured to: determine the type of abnormality based on the abnormal diagnostic results of the audio device; extract abnormal features from the signal features and / or multidimensional sensor data based on the abnormality type, output signal, and / or multidimensional sensor data; determine the degree of abnormality based on the abnormal features and reference output features; and determine graded repair parameters based on the degree of abnormality and the abnormality type.

[0041] In some embodiments, the intelligent diagnostic module is further configured to: generate multiple candidate playback parameters; determine the playback effect of the multiple candidate playback parameters based on abnormal sensing data, signal features, abnormal type and abnormal characteristics, through an effect prediction model; the effect prediction model is a machine learning model; determine the target playback parameter based on the playback effect; and control the audio device to play audio based on the target playback parameter.

[0042] For more information on the above modules, please refer to [link / reference]. Figures 3-5 And related explanations.

[0043] It should be understood that Figure 2 The system and its modules shown can be implemented in various ways. For example, in some embodiments, the various functional modules can be connected to form a closed loop of information operation, and operate in a coordinated and regular manner under the unified management of the intelligent diagnostic module, thereby realizing the informatization and intelligent diagnosis of audio equipment.

[0044] It should be noted that the above description of the system and its modules is for convenience only and should not be construed as limiting this specification to the scope of the illustrated embodiments. It is understood that those skilled in the art, after understanding the principles of the system, may arbitrarily combine the various modules or construct subsystems connected to other modules without departing from these principles. In some embodiments, Figure 2 The sensor module, audio acquisition module, data transmission module, and intelligent diagnostic module disclosed herein can be different modules within a single system, or a single module can perform the functions of two or more of the aforementioned modules. For example, the modules can share a single storage module, or each module can have its own separate storage module. Such variations are all within the scope of protection of this specification.

[0045] Figure 3 This is an exemplary flowchart of a diagnostic method for an audio device according to some embodiments of this specification. Figure 3 As shown, process 300 includes steps 310-340. In some embodiments, process 300 may be executed by diagnostic system 200 or processor 140.

[0046] Step 310: Collect multi-dimensional sensor data from the audio device.

[0047] In some embodiments, step 310 may be performed by the sensor module.

[0048] For more information on audio equipment, please see [link / reference]. Figure 1 And its related descriptions.

[0049] Multidimensional sensor data refers to a collection of data that describes the operating status of an audio device from multiple physical dimensions. For example, multidimensional sensor data can include physical parameters such as temperature, vibration, current, voltage, and spatial orientation.

[0050] In some embodiments, multidimensional sensing data can be acquired using various types of sensors. In some embodiments, different types of sensors can be deployed at different locations on the audio device to acquire multidimensional sensing data from the audio device.

[0051] For example, temperature sensors are placed near heat-generating chips such as power amplifiers in audio equipment to collect temperature data; vibration sensors are installed on mechanical moving parts such as speaker diaphragms in audio equipment to collect vibration data; and current and voltage sensors are connected in series in the power supply circuit of audio equipment to collect current and voltage data.

[0052] Step 320: Acquire the output signal of the audio device and perform feature analysis on the output signal to determine the signal characteristics of the output signal.

[0053] In some embodiments, step 320 may be performed by the audio acquisition module.

[0054] The output signal refers to the audio signal emitted by the audio device during actual playback. For example, the output signal includes not only the sound played normally by the audio device, but also the background noise or abnormal noise generated by the audio device in a specific state (such as standby, fault, etc.).

[0055] In some embodiments, the output signal can be acquired by an audio acquisition module.

[0056] For more information about the audio acquisition module, please refer to [link / reference]. Figure 2 And its related descriptions.

[0057] Feature analysis refers to the signal processing and analysis performed on the acquired output signal. For example, feature analysis includes multi-dimensional analysis such as time-domain feature analysis, frequency-domain feature analysis, and phase feature analysis.

[0058] Signal characteristics refer to quantitative indicators that characterize the state of an output signal. For example, signal characteristics include the frequency distribution, harmonic characteristics, and harmonic distortion of the output signal. Frequency distribution can include the frequencies in the audio signal and the intensity of each frequency. Harmonic characteristics refer to the multiple harmonics in the output signal. Harmonic distortion is the ratio of harmonic energy to the fundamental frequency in the output signal.

[0059] In some embodiments, the audio acquisition module can perform frequency domain feature analysis on the output signal, calculate the frequency distribution, harmonic characteristics and harmonic distortion values ​​of the output signal, and determine them as the signal characteristics of the output signal.

[0060] Step 330: Perform anomaly analysis on the multidimensional sensing data to identify anomalous sensing data.

[0061] In some embodiments, step 330 may be performed by the data transmission module.

[0062] Anomaly analysis refers to the preliminary screening of acquired multidimensional sensor data. For example, anomaly analysis can include threshold detection, pattern recognition, and other multi-faceted analyses.

[0063] In some embodiments, the data transmission module can perform anomaly analysis on the multidimensional sensing data, identifying multidimensional sensing data points that meet a first preset anomaly condition as anomalous data points. The first preset anomaly condition can be that at least one data point in the multidimensional sensing data is greater than a corresponding anomaly threshold. Different data points in the multidimensional sensing data correspond to different anomaly thresholds. The anomaly thresholds can be set empirically.

[0064] For example, the first preset anomaly condition may include at least one of the following: temperature data exceeding a temperature anomaly threshold, vibration data exceeding a vibration anomaly threshold, current data exceeding a current anomaly threshold, and voltage data exceeding a voltage anomaly threshold. Specifically, for spatial attitude data, the first preset anomaly condition may be that the difference between the current spatial attitude data and the average spatial attitude data in historical data exceeds a spatial attitude anomaly threshold. In some embodiments, spatial attitude data can be represented by a vector. The difference between the current spatial attitude data and the average spatial attitude data in historical data can be determined by calculating the vector distance. Methods for calculating the vector distance may include Euclidean distance, cosine distance, etc.

[0065] Abnormal sensor data refers to multidimensional sensor data with labeled abnormal data points.

[0066] In some embodiments, the data transmission module can mark abnormal data points to identify them as abnormal sensing data.

[0067] Step 340: Based on abnormal sensor data and signal characteristics, perform anomaly diagnosis on the audio device.

[0068] In some embodiments, step 340 may be performed by the intelligent diagnostic module.

[0069] Anomaly diagnosis refers to the process of determining the operating status of an audio device and obtaining anomaly diagnosis results. In some embodiments, the anomaly diagnosis results may include abnormal states present in the audio device.

[0070] In some embodiments, the intelligent diagnostic module can perform abnormal diagnostics on the audio device in various ways based on abnormal sensor data and signal characteristics, and determine the abnormal state of the audio device.

[0071] Abnormal conditions can include power amplifier overheating, power amplifier component aging, speaker short circuit, capacitor failure, voice coil scratching, etc.

[0072] In some embodiments, the intelligent diagnostic module can perform anomaly diagnosis on the audio device through cluster analysis based on abnormal sensing data and signal characteristics to determine the abnormal state of the audio device. For example, the intelligent diagnostic module clusters the objects to be clustered, obtaining multiple clusters, wherein the objects to be clustered include multiple cluster vectors and target vectors. The cluster containing the target vector is denoted as the target cluster. Then, the labels of the cluster vectors in the target cluster are calculated and analyzed. The historical abnormal states corresponding to the labels in the target cluster that meet the second preset abnormality condition are determined as the abnormal state of the audio device corresponding to the target vector.

[0073] A cluster vector can include a historical anomaly sensor data point and its corresponding historical signal features. The label corresponding to the cluster vector can be the actual historical anomaly state of the audio device under that historical anomaly sensor data, which can be obtained based on manual diagnosis. The target vector can include the current anomaly sensor data and current signal features of the audio device.

[0074] The second preset anomaly condition may include the historical anomaly state corresponding to the label of the cluster vector, which is the most frequently occurring historical anomaly state corresponding to the label of all cluster vectors in the entire target cluster.

[0075] In some embodiments, the intelligent diagnostic module can perform anomaly diagnosis on the audio device based on abnormal sensor data and signal characteristics, and determine the abnormal state of the audio device. For more detailed explanation, please refer to [link / details]. Figure 4 And its related descriptions.

[0076] In some embodiments of this specification, by simultaneously acquiring multidimensional sensing data and output signals from the audio device, and combining the feature analysis of the output signals with the anomaly analysis of the multidimensional sensing data, a multi-angle and comprehensive diagnosis of the abnormal state of the audio device is achieved, thereby improving the comprehensiveness of the anomaly diagnosis.

[0077] It should be noted that the above description of process 300 is for illustrative purposes only and does not limit the scope of this specification. Those skilled in the art can make various modifications and changes to process 300 under the guidance of this specification. However, these modifications and changes remain within the scope of this specification.

[0078] Figure 4 This is a schematic diagram of an analytical model shown according to some embodiments of this specification.

[0079] In some embodiments, such as Figure 4 As shown, the intelligent diagnostic module is further configured to: determine the abnormal probability distribution 470 based on the abnormal sensing data 411 and signal features 412 through the analysis model 400; and perform abnormal diagnosis on the audio device based on the abnormal probability distribution 470.

[0080] For more information on anomalous sensing data and signal characteristics, please refer to [link / reference]. Figure 3 And its related descriptions.

[0081] An anomaly probability distribution refers to the quantified probability of different abnormal states that an audio device may exhibit. For example, an anomaly probability distribution includes the possible abnormal states of an audio device and the probability corresponding to each abnormal state.

[0082] An analytical model is a model used to determine the probability distribution of anomalies. In some embodiments, the analytical model is a machine learning model. For example, the analytical model may include any one or a combination of deep neural network (DNN) models or other custom model structures.

[0083] In some embodiments, the input to the analysis model may include abnormal sensing data and signal features, and the output may be an abnormal probability distribution.

[0084] In some embodiments, the analysis model can be obtained by training a large number of first training samples with first labels. In some embodiments, a first training sample and its corresponding first label can be determined based on multiple historical playbacks. For example, a first training sample may include the same sample abnormal sensing data and the corresponding same sample signal features from multiple historical playbacks, and the first label may be the actual historical abnormal state and the number of times it occurred when an audio fault occurred in the multiple historical playbacks corresponding to the first training sample. In some embodiments, the first training sample and the first label can be obtained based on historical data.

[0085] In some embodiments, the intelligent diagnostic module can be trained using various methods based on a first training sample and a first label. For example, it can be trained using gradient descent. As an example only, the intelligent diagnostic module can input multiple first training samples with the first label into an initial analysis model, construct a loss function using the first label and the results of the initial analysis model, and iteratively update the parameters of the initial analysis model based on the loss function. When the loss function of the initial analysis model meets preset conditions, the model training is complete, and a trained analysis model is obtained. These preset conditions may include loss function convergence, the number of iterations reaching a threshold, etc.

[0086] In some embodiments, the intelligent diagnostic module can, based on the anomaly probability distribution, identify the anomaly state with the highest probability in the anomaly probability distribution as the current anomaly state of the audio device.

[0087] In some embodiments of this specification, by analyzing the abnormal probability distribution output by the model, the diagnostic results are transformed from a simple "present / absent" abnormal judgment into a quantitative assessment of different abnormal states, thereby achieving a more refined and specific abnormal diagnosis.

[0088] In some embodiments, such as Figure 4As shown, the analysis model 400 includes a feature extraction layer 420 and an anomaly analysis layer 460. The intelligent diagnostic module is further configured to: determine sensing spatial features 431 and signal spatial features 432 through the feature extraction layer 420 based on the abnormal sensing data 411 and signal features 412; and determine the anomaly probability distribution 470 through the anomaly analysis layer 460 based on the sensing spatial features 431 and signal spatial features 432.

[0089] Sensing spatial characteristics refer to features that characterize changes in the physical operating state of an audio device. For example, sensing spatial characteristics include waveform curves, oscillation modes, and multi-axis spatial attitude characteristics of multi-dimensional sensing data. Waveform curves refer to the waveform curves of multi-dimensional sensing data over multiple preset time periods. Oscillation modes refer to the vibration data curves of vibration sensors. Multi-axis spatial attitude characteristics refer to the vibration direction of the audio device and its corresponding oscillation mode.

[0090] Signal spatial characteristics refer to the features that characterize anomalies in the output signal of an audio device. For example, signal spatial characteristics include harmonic structure characteristics, noise characteristics, and frequency band distribution characteristics. Harmonic structure characteristics refer to the spacing and number of harmonics, as well as their energy ratio to the fundamental frequency. Noise characteristics refer to the magnitude and fluctuation range of noise. Frequency band distribution characteristics refer to the energy distribution profile curve of the spectrum in the vertical (i.e., frequency) direction.

[0091] A feature extraction layer is a model structure used to extract features from the sensing space and the signal space. In some embodiments, the feature extraction layer is a machine learning model. For example, the feature extraction layer may include any one or a combination of a Convolutional Neural Network (CNN) model or other custom model structures.

[0092] In some embodiments, the input to the feature extraction layer may include abnormal sensing data and signal features, and the output may be sensing spatial features and signal spatial features.

[0093] In some embodiments, the feature extraction layer can be obtained by training a large number of second training samples with second labels. In some embodiments, the second training samples may include sample anomaly sensing data and sample signal features from historical playback, and the second label may be the actual sensing spatial features and actual signal spatial features from the historical playback corresponding to the second training sample. In some embodiments, the second training samples may be obtained based on historical data, and the second labels corresponding to the second training samples may be obtained by manual annotation.

[0094] The training process of the feature extraction layer is similar to that of the analysis model; please refer to the training process of the analysis model.

[0095] An anomaly analysis layer is a model structure used to determine the probability distribution of anomalies. In some embodiments, the anomaly analysis layer is a machine learning model. For example, the anomaly analysis layer may include any one or a combination of recurrent neural networks or other custom model structures.

[0096] In some embodiments, such as Figure 4 As shown, the analysis model 400 also includes an attention layer 440. The attention layer 440 is configured to determine an attention weight distribution 451 for the sensing spatial features 431 and the signal spatial features 432. The input to the anomaly analysis layer 460 also includes the attention weight distribution 451.

[0097] Attention weight distribution refers to the set of attention weights corresponding to each feature in the sensing space and signal space features when used to determine the probability distribution of anomalies. For example, attention weight distribution can be represented in various forms such as sequences or matrices, where each element corresponds to the attention weight of a feature.

[0098] An attention layer is a model structure used to determine the distribution of attention weights. In some embodiments, the attention layer is a machine learning model. For example, the attention layer may include any one or a combination of recurrent neural networks or other custom model structures.

[0099] In some embodiments, the input to the attention layer may include sensing spatial features and signal spatial features, and the output may be an attention weight distribution.

[0100] In some embodiments, the attention layer can be obtained by training a large number of fourth training samples with fourth labels. In some embodiments, the fourth training samples may include sample sensing space features and sample signal space features, and the fourth label may be the attention weight of each feature in the sample sensing space features and sample signal space features corresponding to the fourth training sample.

[0101] In some embodiments, the fourth training sample can be obtained based on historical data, and the fourth label corresponding to the fourth training sample can be obtained manually. For example, during the labeling process, technicians can judge possible abnormal states (such as amplifier overheating) based on the numerical distribution of each feature in the fourth training sample and their experience, thereby assigning a corresponding attention weight to each feature. It is understandable that since the same type of abnormal state will exhibit similar numerical abnormal distributions (such as a sudden increase in temperature), technicians can mark features related to possible abnormal states as features that need attention and assign higher attention weights to the features that need attention, while reducing the corresponding attention weights for other features that do not need attention.

[0102] The training process for the attention layer is similar to that for the analysis model; please refer to the training process for the analysis model.

[0103] In some embodiments of this specification, by introducing an attention layer to dynamically weight the sensing space features and signal space features, the analysis model can automatically focus on the features that need attention for the current anomaly diagnosis, thereby improving the accuracy of anomaly diagnosis while suppressing noise interference.

[0104] In some embodiments, such as Figure 4 As shown, the input to the anomaly analysis layer 460 also includes reference output features 452.

[0105] In some embodiments, the intelligent diagnostic module is further configured to determine reference output characteristics based on playback parameters and input signal characteristics.

[0106] Playback parameters are those that determine how audio is played. Examples of playback parameters include volume level, equalizer settings, and sound effect modes.

[0107] In some embodiments, playback parameters can be preset by the user on the audio device.

[0108] Input signal characteristics refer to the inherent properties of the audio signal input to an audio device for playback. For example, input signal characteristics include the signal's average energy, dynamic range, and frequency distribution.

[0109] In some embodiments, the input signal characteristics can be determined by analyzing the input audio signal to be played through various methods such as time domain characteristics and frequency domain characteristics.

[0110] Reference output characteristics refer to the range of multidimensional sensor data and the range of output signal characteristics that an audio device should possess during normal operation under different playback parameters and input signal characteristics. For example, reference output characteristics include the temperature range, current range, vibration range, and harmonic distortion range of the audio device's output signal.

[0111] In some embodiments, the intelligent diagnostic module can determine reference output characteristics based on playback parameters and input signal characteristics through various methods. For example, the intelligent diagnostic module can determine the reference output characteristics in a vector database using vector matching based on playback parameters and input signal characteristics.

[0112] The intelligent diagnostic module can construct a matching vector based on the current playback parameters and current input signal characteristics. The vector database includes multiple feature vectors and their corresponding labels. The feature vectors are constructed from multiple historical playback parameters and historical input signal characteristics. The labels corresponding to the feature vectors include the temperature range, current range, vibration range, and harmonic distortion range of the audio device's output signal during multiple normal playbacks, which can be used to detect and acquire information about the audio device.

[0113] In some embodiments, the intelligent diagnostic module can calculate the vector similarity between the vector to be matched and the feature vector, and use the feature vector that meets a first preset condition as the target vector, and use the label corresponding to the target vector as a reference output feature. The first preset condition can be set according to the situation. For example, the highest vector similarity, etc.

[0114] In some embodiments of this specification, by introducing a reference output feature that dynamically changes with playback parameters and input signal characteristics as a benchmark for anomaly diagnosis, the diagnostic system can effectively distinguish between normal parameter fluctuations caused by high load operation and the true abnormal state of the audio device, thereby reducing the false alarm rate.

[0115] In some embodiments, the anomaly analysis layer can be obtained by training a large number of third training samples with third labels. In some embodiments, the third training samples may include sample sensing spatial features and sample signal spatial features during historical playback, and the third label may be the actual historical abnormal state of the audio device and the number of times it occurred among multiple historical audio faults that occurred during historical playback corresponding to the third training sample. In some embodiments, the third training samples may also include sample attention weight distribution and sample reference output features during historical playback.

[0116] In some embodiments, the spatial features of the sample sensing space and the spatial features of the sample signal can be obtained through a trained feature extraction layer.

[0117] In some embodiments, the sample attention weight distribution can be obtained through a trained attention layer.

[0118] In some embodiments, the intelligent diagnostic module can determine the sample reference output characteristics based on the sample playback parameters and sample input signal characteristics during historical playback. For more details, please refer to the relevant description above.

[0119] In some embodiments, the third label may be obtained based on historical data.

[0120] The training process of the anomaly analysis layer is similar to that of the analysis model; please refer to the training process of the analysis model.

[0121] In some embodiments of this specification, the sensing spatial features and signal spatial features in the abnormal sensing data are extracted by the feature extraction layer, and then the time series relationship between the sensing spatial features and signal spatial features is analyzed by the anomaly analysis layer. This enables the analysis model to understand the instantaneous pattern of the abnormal state and its dynamic evolution process at the same time, thereby improving the diagnostic capability for complex time-series abnormal states.

[0122] Figure 5 This is an exemplary flowchart illustrating the determination of graded repair parameters according to some embodiments of this specification.

[0123] In some embodiments, such as Figure 5 As shown, process 500 includes the following steps. In some embodiments, process 500 may be executed by an intelligent diagnostic module or a processor.

[0124] In some embodiments, the diagnostic results for audio device anomalies include the anomaly type. For more information about audio devices, see [link to relevant documentation]. Figure 1 And its related descriptions.

[0125] Anomaly diagnosis results refer to the results obtained by the intelligent diagnostic module after diagnosing anomalies in the audio device. For details on how to perform anomaly diagnosis, please refer to [link / reference needed]. Figure 3 And its related descriptions.

[0126] Anomaly type refers to the category of abnormal states that occur during the operation of audio equipment. Examples include power amplifier overheating, power amplifier component aging, speaker short circuit, capacitor failure, and voice coil scratching.

[0127] Step 510: Extract signal features and / or abnormal features from multidimensional sensing data based on the anomaly type, output signal, and / or multidimensional sensing data.

[0128] For more information on output signals, multidimensional sensor data, and signal characteristics, please refer to [link / reference]. Figure 3 And related content.

[0129] Anomaly features refer to characteristic data used to characterize the abnormal state of audio equipment. In some embodiments, anomaly features can be data directly related to the anomaly type. For example, when the anomaly type is amplifier overheating, the directly related manifestation is a gradual increase in temperature data; in this case, temperature data from multidimensional sensor data is extracted as the anomaly feature. Another example is when the anomaly type is amplifier component aging, the directly related manifestation is a slow increase in harmonic distortion values; in this case, harmonic distortion values ​​from signal features are extracted as the anomaly feature. Yet another example is when the anomaly type is voice coil scratching, the directly related manifestation is irregular high-frequency vibration and increased harmonic distortion values; in this case, vibration data from multidimensional sensor data and harmonic distortion values ​​from signal features are extracted as the anomaly features.

[0130] In some embodiments, anomaly features can be determined by consulting an anomaly feature table based on the anomaly type. The intelligent diagnostic module then extracts data corresponding to signal features and / or multidimensional sensor data based on the determined anomaly features. The anomaly feature table contains the correspondence between anomaly types and anomaly features. In some embodiments, the anomaly feature table can be set based on human experience.

[0131] Step 520: Determine the degree of anomaly based on the anomaly features and the reference output features.

[0132] For more information on reference output characteristics, see [link to relevant documentation]. Figure 4 And its related descriptions.

[0133] Abnormality level refers to an indicator used to measure the degree of deviation of the current abnormal state of an audio device from its normal working state, and is used to reflect the severity of the abnormality.

[0134] In some embodiments, the degree of anomaly can be obtained by calculating the average difference between the value corresponding to the anomaly feature and the corresponding data in the reference output feature. For example, when the anomaly type is amplifier overheating, the anomaly feature corresponds to temperature data, and the degree of anomaly can be the difference between the current temperature data and the closest value of the corresponding temperature range in the reference output feature. As another example, when the anomaly type is voice coil scratching, the anomaly feature corresponds to vibration data and harmonic distortion values. In this case, the degree of anomaly is the average difference between the vibration data, harmonic distortion values, and the closest values ​​of the corresponding vibration range and harmonic distortion range in the reference output feature.

[0135] Step 530: Determine the graded repair parameters based on the degree and type of abnormality.

[0136] Graded repair parameters refer to parameters that trigger different levels of repair or protection measures based on the degree and type of abnormality of the audio device.

[0137] In some embodiments, the graded repair parameters may include primary repair parameters, secondary repair parameters, and tertiary repair parameters. Different repair measures correspond to different levels of repair parameters.

[0138] For illustrative purposes only, the Level 1 repair parameter refers to the protective measure parameter triggered when the anomaly does not affect playback performance. The repair measure corresponding to the Level 1 repair parameter is: the intelligent diagnostic module controls the data transmission module to push a prompt message to the user terminal 120, while simultaneously controlling the audio device to continue normal operation. For more information about the user terminal 120, please refer to [link to relevant documentation]. Figure 1 And its related descriptions.

[0139] Level 2 repair parameters refer to protective measures triggered when an anomaly affects playback performance. The corresponding repair measures are: the intelligent diagnostic module controls the data transmission module to send a warning message to the user terminal 120 and activate the device protection mode. Device protection mode refers to operations taken to prevent device damage. Examples include automatically reducing the maximum volume and temporarily disabling some high-power-consuming functions. The warning message may be "Device overheating, protection mode activated."

[0140] Level 3 repair parameters refer to protective measures triggered when an anomaly may cause equipment damage or pose a safety hazard. The corresponding repair measures for Level 3 repair parameters are: the intelligent diagnostic module controls the user terminal 120 to play a warning voice message and controls the audio device to execute a shutdown procedure. The warning voice message may be "Device is off, please contact after-sales service."

[0141] In some embodiments, the intelligent diagnostic module can determine graded repair parameters based on the degree and type of abnormality using a first preset table.

[0142] The first preset table contains the correspondence between different anomaly types, anomaly severity, and graded repair parameters. The first preset table can be determined based on historical data or prior experience. For example, if the anomaly type is power amplifier overheating, the anomaly severity is an average difference of 2°C between the temperature data and the corresponding data in the reference output characteristics, and the corresponding graded repair parameter is the first-level repair parameter. Another example: if the anomaly type is power amplifier overheating, the anomaly severity is an average difference of 10°C between the temperature data and the corresponding data in the reference output characteristics, and the corresponding graded repair parameter is the third-level repair parameter.

[0143] In some embodiments, the intelligent diagnostic module can determine the current graded repair parameters by querying a first preset table based on the current anomaly type and anomaly severity.

[0144] In some embodiments of this specification, the intelligent diagnostic module extracts abnormal features from the output signal and / or multi-dimensional sensor data based on the anomaly type, and determines the degree of anomaly by combining reference output features, thereby achieving quantitative diagnosis of anomalies. Furthermore, it determines graded repair parameters based on the degree and type of anomaly, enabling the system to automatically match appropriate repair strategies according to the severity of the fault. Through the complete diagnostic process described above, from fault identification, feature extraction, anomaly quantification to graded repair decision-making, intelligent diagnosis from passive detection to proactive control is achieved, improving the accuracy and adaptability of audio equipment fault response.

[0145] In some embodiments, when the graded repair parameters meet preset conditions, the intelligent diagnostic module is further configured to: generate multiple candidate playback parameters; determine the playback effect of the multiple candidate playback parameters based on abnormal sensing data, signal features, abnormal type and abnormal characteristics, through an effect prediction model; the effect prediction model is a machine learning model; determine the target playback parameter based on the playback effect; and control the audio device to play audio based on the target playback parameter.

[0146] Preset conditions refer to the conditions used to determine whether the intelligent diagnostic module should perform operations such as candidate playback parameter generation and effect prediction. In some embodiments, the preset condition may be that the graded repair parameter is a first-level repair parameter.

[0147] Candidate playback parameters refer to a set of playback settings parameters that can be selected. For example, candidate playback parameters could be "volume 20, equalizer bass boost 3, treble boost 2, playback speed 1.0x, maximum volume limit 30".

[0148] In some embodiments, the intelligent diagnostic module can generate candidate playback parameters by performing multiple different random fine-tunings on historical playback parameters.

[0149] The playback effect of candidate playback parameters refers to the overall performance of an audio device when playing audio according to the candidate playback parameters. Playback effect can be characterized in various ways, such as numerical values ​​or levels. As an example, a higher numerical value generally indicates a better playback effect.

[0150] In some embodiments, the effect prediction model refers to a model used to predict the playback effect of candidate playback parameters on an audio device. In some embodiments, the effect prediction model is a machine learning model. For example, the effect prediction model may include one or more combinations of Deep Neural Network (DNN) models, Convolutional Neural Network (CNN) models, or other custom models.

[0151] In some embodiments, the input to the effect prediction model may be anomaly sensing data, signal features, anomaly type, and candidate playback parameters, and the output of the effect prediction model may be the playback effect corresponding to the candidate playback parameters. For more information on anomaly sensing data, see [link to relevant documentation]. Figure 3 And its related descriptions.

[0152] In some embodiments, the effect prediction model can be obtained by training an initial effect prediction model using multiple sets of fifth training samples with fifth labels. The fifth training samples may include historical anomaly sensor data, historical signal features, historical anomaly types, and historical playback parameters from previous playback sessions. The fifth label can be the actual playback effect of the historical playback corresponding to the fifth training sample. For more information on historical anomaly sensor data and historical signal features, see [link to relevant documentation]. Figure 3 And related descriptions. In some embodiments, historical playback parameters can be determined by querying user settings in historical data. In some embodiments, historical anomaly types can be determined based on manual diagnosis.

[0153] In some embodiments, the fifth tag can be obtained by weighted summation of the normalized values ​​of audio effects and device load during historical playback, as shown in equation (1) below: .

[0154] In equation (1), Indicates the playback effect; This indicates the normalized audio effect. This represents the normalized device load. and These represent the weighting coefficients for audio effects and device load, respectively. It is a negative number.

[0155] In some embodiments, and The value can be set by technicians based on experience.

[0156] Normalization refers to converting indicators with different dimensions or value ranges to the same standard range.

[0157] In some embodiments, audio effects It can be represented by the waveform similarity between the input and output signals and the noise level of the output signal. In some embodiments, the higher the waveform similarity between the input and output signals and the lower the noise of the output signal, the better the audio playback effect.

[0158] In some embodiments, when the device load and the audio device play audio according to candidate playback parameters, the values ​​corresponding to the multidimensional sensing data collected by the sensor module are positively correlated with the difference between the values ​​and the normal ranges of the data from various sensors. The normal ranges of the data from various sensors can be preset by technicians based on experience.

[0159] In some embodiments, the fifth training sample and the fifth label may be obtained based on historical data.

[0160] In some embodiments, the training process of the effect prediction model is similar to that of the analysis model; see the training process of the analysis model for details.

[0161] The target playback parameters refer to the final playback parameters determined after the playback effect is predicted.

[0162] In some embodiments, the target playback parameter is the candidate playback parameter with the largest playback effect output by the effect prediction model among a plurality of candidate playback parameters.

[0163] In some embodiments, the intelligent diagnostic module sends the target playback parameters to the audio device through the data transmission module and controls the audio device to play audio based on the target playback parameters.

[0164] In some embodiments of this specification, when the graded repair parameters meet preset conditions, the intelligent diagnostic module uses an effect prediction model to evaluate the playback effect of multiple candidate playback parameters and determines the target playback parameter based on the prediction results, thereby achieving intelligent adaptive optimization of the audio device's playback parameters. By comprehensively analyzing abnormal sensor data, signal characteristics, abnormal types, and abnormal features through a machine learning model, the module dynamically adjusts device operating parameters without interrupting the user's playback experience, achieving real-time compensation and trend suppression for minor abnormalities, thus improving the system's online self-repair capability and operational stability.

[0165] It should be noted that the above description of process 500 is merely for illustration and explanation, and does not limit the scope of this specification. Those skilled in the art can make various modifications and changes to process 500 under the guidance of this specification. However, these modifications and changes are still within the scope of this specification. The basic concepts have been described above; obviously, those skilled in the art will find that the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification, and therefore, such modifications, improvements, and corrections still fall within the spirit and scope of the exemplary embodiments of this specification.

[0166] Furthermore, this specification uses specific terms to describe embodiments thereof. For example, "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Moreover, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined.

[0167] Furthermore, unless expressly stated in the claims, the order of processing elements and sequences, the use of numbers and letters, or other names described in this specification are not intended to limit the order of the processes and methods described herein. Although various examples have been discussed in the foregoing disclosure of some embodiments of the invention that are currently considered useful, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments; rather, the claims are intended to cover all modifications and equivalent combinations that conform to the spirit and scope of the embodiments described herein. For example, while the system components described above can be implemented using hardware devices, they can also be implemented solely using software solutions, such as installing the described system on existing servers or mobile devices.

[0168] Similarly, it should be noted that, in order to simplify the description disclosed herein and thus aid in the understanding of one or more embodiments of the invention, the foregoing description of embodiments in this specification may sometimes combine multiple features into a single embodiment, drawing, or description thereof. However, this method of disclosure does not imply that the subject matter of this specification requires more features than those mentioned in the claims. In fact, the embodiments contain fewer features than all the features of a single embodiment disclosed above.

[0169] In some embodiments, numbers describing the quantity of components and attributes are used. It should be understood that such numbers used in the description of embodiments are modified in some examples with the terms "approximately," "approximately," or "generally." Unless otherwise stated, "approximately," "approximately," or "generally" indicates that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may be changed depending on the characteristics required by individual embodiments. In some embodiments, numerical parameters should take into account specified significant digits and employ a general method of digit reservation. Although the numerical ranges and parameters used to confirm their breadth of range in some embodiments of this specification are approximate values, in specific embodiments, such values ​​are set as precisely as feasible.

[0170] For each patent, patent application, patent application publication, and other material such as articles, books, specifications, publications, and documents referenced in this specification, the entire contents of which are incorporated herein by reference. This excludes historical application documents that are inconsistent with or conflict with the content of this specification, as well as documents that limit the broadest scope of the claims in this specification (currently or subsequently appended to this specification). It should be noted that in the event of any inconsistency or conflict between the descriptions, definitions, and / or terminology used in the supplementary materials to this specification and the content of this specification, the descriptions, definitions, and / or terminology used in this specification shall prevail.

[0171] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and are considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A diagnostic system for audio equipment, characterized in that, The diagnostic system includes: The sensor module is configured to acquire multidimensional sensing data from the audio device; The audio acquisition module is configured to: acquire the output signal of the audio device, perform feature analysis on the output signal, and determine the signal characteristics of the output signal; The data transmission module is configured to: perform anomaly analysis on the multidimensional sensing data, determine the abnormal sensing data, and transmit the signal features and the abnormal sensing data to the intelligent diagnostic module; The intelligent diagnostic module is configured to perform anomaly diagnosis on the audio device based on the abnormal sensing data and the signal characteristics.

2. The diagnostic system as described in claim 1, characterized in that, The intelligent diagnostic module is further configured as follows: Based on the abnormal sensing data and the signal characteristics, the anomaly probability distribution is determined through an analysis model; the analysis model is a machine learning model. The abnormality diagnosis is performed on the audio device based on the aforementioned abnormality probability distribution.

3. The diagnostic system as described in claim 2, characterized in that, The analysis model includes a feature extraction layer and an anomaly analysis layer; the intelligent diagnostic module is further configured as follows: Based on the abnormal sensing data and the signal features, the sensing spatial features and signal spatial features are determined through the feature extraction layer. Based on the sensing spatial features and the signal spatial features, the anomaly probability distribution is determined through the anomaly analysis layer.

4. The diagnostic system as described in claim 3, characterized in that, The input to the anomaly analysis layer also includes reference output features, and the intelligent diagnostic module is further configured as follows: The reference output characteristics are determined based on the playback parameters and input signal characteristics.

5. The diagnostic system as described in claim 1, characterized in that, The abnormality diagnosis results of the audio device include the abnormality type, and the intelligent diagnosis module is further configured as follows: Based on the anomaly type, the output signal, and / or the multidimensional sensing data, extract the signal features and / or the anomaly features in the multidimensional sensing data; The degree of anomaly is determined based on the aforementioned anomaly characteristics and reference output characteristics; Based on the degree and type of the anomaly, the graded repair parameters are determined.

6. A diagnostic method for an audio device, characterized in that, The diagnostic method includes: Collect multidimensional sensor data from the audio device; The output signal of the audio device is acquired, and feature analysis is performed on the output signal to determine the signal characteristics of the output signal; Anomaly analysis is performed on the multidimensional sensing data to identify abnormal sensing data; Based on the abnormal sensing data and the signal characteristics, the audio device is diagnosed as abnormal.

7. The diagnostic method as described in claim 6, characterized in that, The abnormality diagnosis of the audio device based on the abnormal sensing data and the signal characteristics includes: Based on the abnormal sensing data and the signal characteristics, the anomaly probability distribution is determined through an analysis model; the analysis model is a machine learning model. The abnormality diagnosis is performed on the audio device based on the aforementioned abnormality probability distribution.

8. The diagnostic method as described in claim 7, characterized in that, The analysis model includes a feature extraction layer and an anomaly analysis layer. The step of determining the anomaly probability distribution based on the anomaly sensing data and the signal features through the analysis model includes: Based on the abnormal sensing data and the signal features, the sensing spatial features and signal spatial features are determined through the feature extraction layer. Based on the sensing spatial features and the signal spatial features, the anomaly probability distribution is determined through the anomaly analysis layer.

9. The diagnostic method as described in claim 8, characterized in that, The input to the anomaly analysis layer also includes reference output features, and the diagnostic method further includes: The reference output characteristics are determined based on the playback parameters and input signal characteristics.

10. The diagnostic method as described in claim 6, characterized in that, The abnormality diagnosis results of the audio device include the abnormality type, and the diagnostic method further includes: Based on the anomaly type, the output signal, and / or the multidimensional sensing data, extract the signal features and / or the anomaly features in the multidimensional sensing data; The degree of anomaly is determined based on the aforementioned anomaly characteristics and reference output characteristics; Based on the degree and type of the anomaly, the graded repair parameters are determined.

Images