Industrial diagnosis method based on voiceprint recognition and related equipment

Through the method based on voiceprint recognition, the sound data of industrial equipment is collected and analyzed, and the deep neural network model is established, contactless fault diagnosis is realized, the accuracy and inefficiency of image analysis methods are solved, and industrial diagnosis is improved Accuracy and efficiency.

CN119068912APending Publication Date: 2024-12-03PING AN TECH (SHENZHEN) CO LTD

Patent Information

Application Number
CN202411246781.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-05
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

In industrial diagnosis, the image analysis method has problems such as equipment being easily damaged, high cost and environmentally affected by the image quality, resulting in insufficient diagnostic accuracy and inefficient efficiency.

Method used

Using a method based on voiceprint recognition, the sound data of industrial equipment in normal and faulty states is collected, preprocessing and feature extraction is performed, a deep neural network model is established, and voiceprint feature comparison is performed to realize contactless fault diagnosis.

Benefits of technology

提高了工业诊断的准确性和效率,降低了成本,增强了诊断的灵活性和可靠性。

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119068912A_ABST
    Figure CN119068912A_ABST
Patent Text Reader

Abstract

The embodiment of the invention belongs to the field of industrial diagnosis and the field of financial science and technology, and relates to a voiceprint recognition-based industrial diagnosis method, which comprises the following steps of: acquiring sound data of industrial equipment running in a normal state and a fault state; performing preprocessing and feature extraction on the sound data to obtain voiceprint feature vectors; establishing a voiceprint model through a deep neural network by using the voiceprint feature vector; when target sound data of to-be-diagnosed industrial equipment is received, performing feature comparison on the target sound data according to the voiceprint model to obtain a feature comparison result; and performing diagnosis analysis according to the feature comparison result to obtain a diagnosis result of the to-be-diagnosed industrial equipment. The invention further provides an industrial diagnosis device based on voiceprint recognition, computer equipment and a storage medium. According to the method and the device, voiceprint recognition is automatically performed on the sound data generated by the industrial equipment, so that the accuracy of industrial diagnosis is improved, and the efficiency of industrial diagnosis is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the fields of industrial diagnosis and financial technology, and is applied to the scenario of fault diagnosis for industrial equipment, and in particular to an industrial diagnosis method and related equipment based on voiceprint recognition. Background Art

[0002] As the pillar of the national economy, industry has promoted economic growth through technological innovation and productivity improvement. In the industrial production environment, industrial diagnosis is crucial to ensure the stable operation of industrial equipment, which can prevent failures and improve production efficiency.

[0003] In recent years, the means of industrial diagnosis have gradually developed towards intelligence. For example, image analysis methods can determine whether there are problems with industrial equipment by analyzing images taken by cameras and other devices. Compared with traditional manual troubleshooting methods, this not only reduces time and labor costs, but also makes industrial diagnosis more flexible and less susceptible to limitations.

[0004] However, the image analysis method faces the problem that the equipment used for diagnosis is easily damaged, resulting in high costs, and the image quality is easily affected by environmental factors. Therefore, it is necessary to propose a solution that aims to reduce the cost of industrial diagnosis and improve the accuracy and efficiency of industrial diagnosis.

[0005] In addition, this solution is not limited to the industrial field, but can also be extended to fields such as financial technology, supporting decision-making through precise data analysis to improve the security and reliability of financial services. Summary of the invention

[0006] The purpose of the embodiments of the present application is to propose an industrial diagnosis method and related equipment based on voiceprint recognition to solve the technical problems of insufficient accuracy and low efficiency of industrial diagnosis.

[0007] In order to solve the above technical problems, the embodiment of the present application provides an industrial diagnosis method based on voiceprint recognition, which adopts the following technical solution:

[0008] An industrial diagnosis method based on voiceprint recognition comprises the following steps:

[0009] Collect sound data of industrial equipment operating in normal and fault conditions;

[0010] Preprocessing and feature extracting the sound data to obtain a voiceprint feature vector;

[0011] Using the voiceprint feature vector, establishing a voiceprint model through a deep neural network;

[0012] When receiving target sound data of the industrial equipment to be diagnosed, performing feature comparison on the target sound data according to the voiceprint model to obtain a feature comparison result;

[0013] A diagnostic analysis is performed based on the feature comparison result to obtain a diagnostic result of the industrial equipment to be diagnosed.

[0014] Furthermore, when the target sound data of the industrial equipment to be diagnosed is received, the step of performing feature comparison on the target sound data according to the voiceprint model to obtain the feature comparison result specifically includes:

[0015] When the target sound data is received, preprocessing and feature extraction are performed on the target sound data to obtain a target voiceprint feature vector;

[0016] The target voiceprint feature vector is input into the voiceprint model, and the similarity between the target voiceprint feature vector and the voiceprint feature vector is calculated as the feature comparison result.

[0017] Furthermore, the step of inputting the target voiceprint feature vector into the voiceprint model and calculating the similarity between the target voiceprint feature vector and the voiceprint feature vector as the feature comparison result specifically includes:

[0018] Using metric learning technology, establishing a metric learning model according to the voiceprint feature vector;

[0019] The similarity is calculated based on the metric learning model.

[0020] Furthermore, the step of performing diagnostic analysis based on the feature comparison result to obtain the diagnostic result of the industrial equipment to be diagnosed specifically includes:

[0021] Comparing the similarity with a preset threshold to obtain a comparison result;

[0022] Analyze according to the comparison result to determine the device state of the industrial device to be diagnosed, wherein the device state includes a normal state and a fault state;

[0023] If the device state is a fault state, determining the corresponding fault type;

[0024] A diagnosis analysis report is generated according to the comparison result, the device status and the fault type as the diagnosis result.

[0025] Furthermore, the step of preprocessing and extracting features from the sound data to obtain a voiceprint feature vector specifically includes:

[0026] De-noising, filtering and enhancing signal quality of the sound data to obtain pre-processed training data;

[0027] The training data is subjected to feature extraction according to a Mel-frequency cepstral coefficient algorithm to obtain the voiceprint feature vector.

[0028] Furthermore, the step of extracting features from the training data according to the Mel-frequency cepstral coefficient algorithm to obtain the voiceprint feature vector specifically includes:

[0029] Performing frame division, windowing and Fourier transformation on the training data to obtain a spectrum signal;

[0030] Filtering the spectrum signal through a Mel filter bank to obtain an energy response;

[0031] Performing logarithmic operation and cepstral coefficient extraction on the energy response to obtain cepstral coefficients;

[0032] The voiceprint feature vector is formed according to the cepstral coefficients.

[0033] Furthermore, after the step of performing diagnostic analysis according to the feature comparison result to obtain the diagnostic result of the industrial equipment to be diagnosed, the method further includes:

[0034] Receiving a data stream corresponding to the sound data in real time, and extracting incremental features;

[0035] An online learning algorithm is used to perform incremental training on the voiceprint model according to the incremental features to achieve dynamic updating of the voiceprint model.

[0036] In order to solve the above technical problems, the embodiment of the present application also provides an industrial diagnosis device based on voiceprint recognition, which adopts the following technical solution:

[0037] An industrial diagnostic device based on voiceprint recognition, comprising:

[0038] A collection module is used to collect sound data of industrial equipment operating in normal and fault conditions;

[0039] An extraction module, used for preprocessing and feature extraction of the sound data to obtain a voiceprint feature vector;

[0040] A building module, used to use the voiceprint feature vector to build a voiceprint model through a deep neural network;

[0041] A comparison module, configured to, when receiving target sound data of the industrial equipment to be diagnosed, perform feature comparison on the target sound data according to the voiceprint model to obtain a feature comparison result;

[0042] The analysis module is used to perform diagnostic analysis based on the feature comparison result to obtain a diagnostic result of the industrial equipment to be diagnosed.

[0043] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0044] A computer device includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the industrial diagnosis method based on voiceprint recognition as described above when executing the computer-readable instructions.

[0045] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0046] A computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the industrial diagnosis method based on voiceprint recognition as described above.

[0047] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0048] The industrial diagnosis method based on voiceprint recognition disclosed in the present application collects sound data of industrial equipment running in normal state and fault state; and pre-processes and extracts features of the sound data to obtain voiceprint feature vectors; then, uses the voiceprint feature vectors to establish a voiceprint model through a deep neural network; when the target sound data of the industrial equipment to be diagnosed is received, the target sound data is feature-matched according to the voiceprint model to obtain feature comparison results; and then a diagnostic analysis is performed based on the feature comparison results to obtain the diagnostic results of the industrial equipment to be diagnosed. By establishing voiceprint models of industrial equipment in different states, the present application can automatically perform voiceprint recognition on the sound data generated by the industrial equipment to achieve non-contact fault diagnosis, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0050] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0051] Figure 2is a flow chart of an embodiment of an industrial diagnosis method based on voiceprint recognition according to the present application;

[0052] Figure 3 is a structural schematic diagram of an embodiment of an industrial diagnostic device based on voiceprint recognition according to the present application;

[0053] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0054] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0055] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0056] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0057] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0058] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0059] The terminal devices 101, 102, and 103 can be various electronic devices with display screens and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 (Moving Picture Experts Group Audio Layer III) players, MP4 (Moving Picture Experts Group Audio Layer IV) players, laptop computers, desktop computers, and the like.

[0060] The server 105 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal devices 101 , 102 , and 103 .

[0061] It should be noted that the industrial diagnosis method based on voiceprint recognition provided in the embodiment of the present application is generally executed by a server, and accordingly, the industrial diagnosis device based on voiceprint recognition is generally set in the server.

[0062] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0063] Continue to refer Figure 2 , shows a flow chart of an embodiment of an industrial diagnosis method based on voiceprint recognition according to the present application. The industrial diagnosis method based on voiceprint recognition comprises the following steps:

[0064] Step S201 : collecting sound data of industrial equipment operating in a normal state and a fault state.

[0065] It should be noted that the relevant technical solutions of voiceprint recognition in this application can also be applied to the field of financial technology. For example, it can be applied to intelligent customer service systems to identify customer emotions or verify customer identities by analyzing the user's voice characteristics, thereby improving customer service experience and security. In addition, voiceprint recognition can also be used to monitor abnormal transaction behaviors and assist in identifying potential fraudulent behaviors by real-time monitoring of changes in user voice patterns. In short, voiceprint recognition has great application potential in the field of financial technology. It can support decision-making through accurate data analysis as an effective means to improve service quality and risk control level.

[0066] In this embodiment, the industrial diagnosis method based on voiceprint recognition is executed on the electronic device (eg Figure 1The server shown in the figure) can send or receive data through a wired connection or a wireless connection. It should be noted that the above wireless connection method may include but is not limited to 3G / 4G / 5G connection, Wi-Fi connection, Bluetooth connection, Wi MAX connection, Zigbee connection, UWB (Ultra Wi-Fi Band) connection, and other wireless connection methods currently known or to be developed in the future.

[0067] Specifically, in the process of collecting sound data, the industrial equipment to be collected can be selected from commonly used equipment in the industrial field, such as a motor, and the environment needs to be relatively quiet to ensure that there is no interference sound generated by other mechanical equipment around the industrial equipment to reduce the impact of background noise. At the same time, microphone arrays or sensors can be selected to collect sound data. Placing highly sensitive microphone arrays or sensors near the industrial equipment can minimize the attenuation of sound waves during propagation. In addition, since it is necessary to collect sound data of industrial equipment running in normal state and fault state, it can be divided into two collection processes. For example, the process of collecting sound data of industrial equipment running in normal state includes: putting the equipment into normal state for operation, ensuring that the equipment is fully preheated to a stable operating temperature, and using a microphone array or sensor to continuously record the sound of the equipment when it is running. The recording time is 5 minutes, and the recording process is repeated to obtain at least 10 sets of data; the process of collecting sound data of industrial equipment running in fault state includes: artificially simulating various fault types, such as adjusting certain components of the equipment to deviate from the normal working state, and recording the sound data of the equipment running under different fault types. Each fault type is recorded for 5 minutes. Similarly, at least 10 sets of data are collected for each fault type; using this method to collect sound data can ensure the diversity and sufficiency of the data. Finally, each set of data is marked, including information such as equipment status (normal state or fault state), fault type (if any), collection time, etc., and is organized into a unified format for storage to facilitate subsequent preprocessing and feature extraction.

[0068] Step S202: preprocessing and feature extraction are performed on the sound data to obtain a voiceprint feature vector.

[0069] Specifically, the collected sound data first needs to be preprocessed to ensure the accuracy of subsequent analysis, which mainly includes three processes: noise removal, filtering, and signal quality enhancement. Noise removal can remove high-frequency noise in the sound data and retain the signal in the main frequency range; filtering can convert the signal to the frequency domain and remove noise; and signal quality enhancement can improve the clarity of the signal. Then, feature extraction can be performed on the preprocessed sound data, and effective voiceprint features can be extracted from the preprocessed sound data to form a voiceprint feature vector for subsequent establishment of a voiceprint model. The means of feature extraction include spectrogram, Mel-frequency cepstral coefficients (MFCCs), short-time energy, etc. In this embodiment, Mel-frequency cepstral coefficients are mainly used for feature extraction.

[0070] It should be noted that Mel Frequency Cepstrum Coefficients (MFCCs) is a feature extraction method widely used in the fields of speech recognition and voiceprint recognition. It can capture the spectral characteristics of sound signals and simulate the perception characteristics of the human ear. Therefore, it is very suitable for voiceprint recognition tasks. In the following embodiments, the process of preprocessing and feature extraction will be further explained and will not be repeated here.

[0071] Step S203: Using the voiceprint feature vector, a voiceprint model is established through a deep neural network.

[0072] Specifically, the voiceprint model belongs to a deep neural network model and can be established using a machine learning algorithm. For example, the voiceprint feature vector can first be divided into a training set, a validation set, and a test set, with a ratio of approximately 80%, 10%, and 10%, to ensure that each set contains voiceprint feature vectors in a normal state and a fault state, and the voiceprint feature vectors in the training set are standardized to ensure that they are on the same scale. Then, a suitable deep neural network architecture needs to be selected, such as a convolutional neural network (CNN), a recurrent neural network (RNN), or a long short-term memory network (LSTM). In this embodiment, a recurrent neural network (RNN) that is more suitable for voiceprint recognition is selected. The network can capture the timing and semantic information in the voiceprint through multi-layer stacking, attention mechanism, and timing modeling. After the initial model is established, the training set can be used for model training. During the training process, since the collected sound data is labeled, the weight parameters of the model can be updated through the back propagation algorithm, and the trained model can be evaluated and tuned using the validation set and the test set. The evaluation indicators may include the equal error rate (Equa l Er ror The tuning contents include model architecture, hyperparameters or data, so as to improve the model performance and finally obtain the established voiceprint model.

[0073] It is understandable that when the voiceprint model is applied to actual scenarios, the deployment of the voiceprint model can be local deployment on an embedded device or cloud deployment that provides services through a network interface. After inputting a new voice sample, the voiceprint model can perform feature extraction and voiceprint matching, thereby realizing the function of voiceprint recognition.

[0074] Step S204, when the target sound data of the industrial equipment to be diagnosed is received, feature comparison is performed on the target sound data according to the voiceprint model to obtain a feature comparison result.

[0075] Specifically, after the voiceprint model is deployed, the target sound data of the industrial equipment to be diagnosed can be compared with its features through the voiceprint model, and industrial diagnosis can be achieved by voiceprint recognition. For example, the target sound data is also preprocessed, including noise removal, filtering and signal quality enhancement, and then the Mel-frequency cepstral coefficients are used for feature extraction to obtain the target voiceprint feature vector. The target voiceprint feature vector is input into the deployed voiceprint model, and the voiceprint model is used to calculate the similarity between the target voiceprint feature vector and the voiceprint feature vector stored in the voiceprint model as the feature comparison result. Subsequently, the equipment status of the industrial equipment to be diagnosed can be judged according to the feature comparison result, and industrial diagnosis can be performed on the industrial equipment to be diagnosed.

[0076] Step S205 , performing diagnostic analysis based on the feature comparison result to obtain a diagnostic result of the industrial equipment to be diagnosed.

[0077] Specifically, the feature comparison result includes the similarity between the target voiceprint feature vector and the voiceprint feature vector stored in the voiceprint model. Since the voiceprint feature vector stored in the voiceprint model is marked, it can correspond to different equipment states (normal state or fault state). If the equipment state is a fault state, it can also correspond to different fault types. Therefore, multiple thresholds can be set in advance, and each threshold corresponds to a different fault type. When performing diagnostic analysis, the similarity between the target voiceprint feature vector and the voiceprint feature vector stored in the voiceprint model can be compared with the preset threshold. If the similarity between the target voiceprint feature vector and the voiceprint feature vector corresponding to a certain fault type is lower than the preset threshold corresponding to the fault type, it can be judged that the industrial equipment to be diagnosed is in a fault state and there is a fault type corresponding to the preset threshold. If the similarity between the target voiceprint feature vector and the voiceprint feature vector corresponding to each fault type exceeds or is equal to each preset threshold, it can be judged that the industrial equipment to be diagnosed is in a normal state. Finally, a diagnostic analysis report can be generated according to the results of the above judgment as the diagnostic result of the industrial equipment to be diagnosed.

[0078] It is understandable that each fault type can also be associated with one or more threshold intervals. When performing diagnostic analysis, if the industrial equipment to be diagnosed has a fault type corresponding to a certain preset threshold, the degree of the fault type can be further judged based on the threshold interval in which the similarity lies, so as to obtain a more detailed diagnostic result.

[0079] By establishing voiceprint models of industrial equipment in different states, the present application can automatically perform voiceprint recognition on sound data generated by industrial equipment to achieve non-contact fault diagnosis, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0080] In some optional implementations of this embodiment, when receiving the target sound data of the industrial equipment to be diagnosed, the step of performing feature comparison on the target sound data according to the voiceprint model to obtain the feature comparison result includes:

[0081] When the target sound data is received, preprocessing and feature extraction are performed on the target sound data to obtain a target voiceprint feature vector;

[0082] The target voiceprint feature vector is input into the voiceprint model, and the similarity between the target voiceprint feature vector and the voiceprint feature vector is calculated as the feature comparison result.

[0083] In this embodiment, when the target sound data of the industrial equipment to be diagnosed is received, the received target sound data needs to be preprocessed and feature extracted in the same manner as the sound data processing to obtain the target voiceprint feature vector, and then the target voiceprint feature vector can be input into the voiceprint model and feature-compared with the voiceprint feature vector stored in the voiceprint model. Specifically, the metric learning technology can be used to calculate the similarity between the target voiceprint feature vector and the voiceprint feature vector stored in the voiceprint model. The metric learning technology includes contrast loss and triplet loss. The contrast loss is applicable to the similarity calculation between a pair of feature vectors, and its goal is to minimize the distance between samples of the same type and maximize the distance between samples of different types; while the triplet loss is applicable to the similarity calculation between triplet feature vectors. The triplet consists of an anchor feature vector, a positive sample feature vector and a negative sample feature vector, and its goal is to shorten the distance between the anchor point and the positive sample and increase the distance between the anchor point and the negative sample. By using contrast loss or triplet loss, the target voiceprint feature vector can be compared with the voiceprint feature vector stored in the voiceprint model. By calculating the similarity between the two, it is determined whether the two feature vectors match, so that the similarity can be used for subsequent diagnostic analysis to achieve industrial diagnosis.

[0084] The present application ensures the accuracy of feature comparison by calculating the similarity between feature vectors and provides a data basis for subsequent diagnostic analysis, so that non-contact fault diagnosis can be achieved based on the similarity, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0085] In some optional implementations of this embodiment, the step of inputting the target voiceprint feature vector into the voiceprint model, calculating the similarity between the target voiceprint feature vector and the voiceprint feature vector as the feature comparison result includes:

[0086] Using metric learning technology, establishing a metric learning model according to the voiceprint feature vector;

[0087] The similarity is calculated based on the metric learning model.

[0088] In this embodiment, contrast loss or triplet loss can be selected as the metric learning technology, and the metric learning model is established according to the voiceprint feature vector stored in the voiceprint model. Specifically, the metric learning model can be regarded as an enhanced component of the voiceprint model, which is used to learn the distance metric function to improve the accuracy and robustness of the voiceprint model and optimize the performance of the voiceprint model. For example, contrast loss is selected as the metric learning technology, a metric learning model is established, and the Euclidean distance or cosine similarity between each two feature vectors is calculated according to the voiceprint feature vector stored in the voiceprint model, and then the loss value is calculated according to the contrast loss formula, and the loss function of the metric learning model is defined to complete the model training, so that the trained metric learning model can be used to perform distance measurement on the target voiceprint feature vector, that is, the Euclidean distance between the target voiceprint feature vector and the voiceprint feature vector stored in the voiceprint model is calculated as the similarity.

[0089] It should be noted that contrast loss is a commonly used metric learning loss function, which is used to learn the distance between similar samples and dissimilar samples in supervised learning tasks. The goal of the contrast loss function is to minimize the distance between samples of the same type and maximize the distance between samples of different types. The contrast loss function for calculating the Euclidean distance can be expressed as:

[0090]

[0091] Where N is the total number of samples, d(x i ,x j ) is the sample x i and x j The distance between ijIt is a label indicating whether two samples belong to the same category, and m is a margin value used to control the minimum distance between different-class samples.

[0092] This application calculates the similarity between the target voiceprint feature vector and the voiceprint feature vectors stored in the voiceprint model through metric learning technology, thereby obtaining a more accurate feature comparison result, providing a data basis for subsequent diagnostic analysis, and enabling non-contact fault diagnosis based on the similarity, increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0093] In some alternative implementation manners of this embodiment, the step of performing diagnostic analysis based on the feature comparison result to obtain the diagnostic result of the industrial device to be diagnosed includes:

[0094] Compare the similarity with a preset threshold to obtain a comparison result;

[0095] Analyze according to the comparison result to determine the device state of the industrial device to be diagnosed, where the device state includes a normal state and a fault state;

[0096] If the device state is a fault state, then determine the corresponding fault type;

[0097] Generate a diagnostic analysis report based on the comparison result, the device state, and the fault type as the diagnostic result.

[0098] In this embodiment, multiple thresholds can be preset based on historical experimental results and the requirements of the actual application scenario, and each threshold corresponds to a different type of fault. Specifically, common fault types include bearing wear, motor overheating, mechanical component looseness, etc. Set a threshold for each fault type. Since the voiceprint feature vectors corresponding to each fault type are stored in the voiceprint model, when obtaining the similarities between the target voiceprint vector and the feature vectors stored in the voiceprint model through feature comparison, each similarity can be compared with the preset threshold of the corresponding fault type to determine whether there is a similarity lower than the preset threshold. If so, it is determined that the industrial device to be diagnosed has this type of fault; if all similarities exceed or are equal to the preset threshold, it is determined that the device state of the industrial device to be diagnosed is a normal state. In addition, one or more threshold intervals can be set for each fault type to represent different degrees of faults. For example, for bearing wear, two thresholds T1 and T2 (T1 < T2) can be set, where: if the value of the similarity is in the range [T1, T2), it is diagnosed as mild bearing wear; if the similarity value is in the range [T2, 1], it is diagnosed as severe bearing wear.

[0099] It is understandable that if the similarity is close to a preset threshold, or fluctuates around the preset threshold, further inspection may be required. At this time, the detection device to be diagnosed can be marked as an uncertain state and used as a diagnosis result.

[0100] The present application can determine the equipment status and fault type of industrial equipment through diagnostic analysis of feature comparison results, thereby obtaining diagnostic results, realizing non-contact fault diagnosis, increasing the accuracy of industrial diagnosis, and improving the efficiency of industrial diagnosis.

[0101] In some optional implementations of this embodiment, the steps of preprocessing and extracting features from the sound data to obtain a voiceprint feature vector include:

[0102] De-noising, filtering and enhancing signal quality of the sound data to obtain pre-processed training data;

[0103] The training data is subjected to feature extraction according to a Mel-frequency cepstral coefficient algorithm to obtain the voiceprint feature vector.

[0104] In this embodiment, the preprocessing process includes removing noise, filtering and enhancing signal quality. After completing the preprocessing of the sound data to obtain the training data, the training data can be feature extracted according to the Mel-frequency cepstral coefficient algorithm to obtain the voiceprint feature vector. Specifically, the sound data is preprocessed. First, a bandpass filter can be used to remove high-frequency noise in the target sound data to retain the signal within the main frequency range, and then the signal is converted to the frequency domain using the FFT transform, and the filter is applied to remove noise. Then, the gain of the target sound data is adjusted to improve the clarity of the signal. In addition, to ensure the consistency of the data, downsampling, normalization and other processes can be performed on the sound data, or dynamic range compression technology can be used to balance the loudness of the signal as needed.

[0105] This application ensures the quality of the feature vector by preprocessing the sound data and using the Mel-frequency cepstral coefficient algorithm for feature extraction, provides a basis for the subsequent establishment of the voiceprint model, and improves the robustness and generalization ability of the voiceprint model. Therefore, non-contact fault diagnosis can be achieved with the help of the voiceprint model, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0106] In some optional implementations of this embodiment, the step of extracting features from the training data according to the Mel-frequency cepstral coefficient algorithm to obtain the voiceprint feature vector includes:

[0107] Performing frame division, windowing and Fourier transformation on the training data to obtain a spectrum signal;

[0108] Filtering the spectrum signal through a Mel filter bank to obtain an energy response;

[0109] Performing logarithmic operation and cepstral coefficient extraction on the energy response to obtain cepstral coefficients;

[0110] The voiceprint feature vector is formed according to the cepstral coefficients.

[0111] In this embodiment, the specific process of the Mel-frequency cepstral coefficient algorithm may include framing, windowing, Fourier transform, Mel filter group, logarithmic operation, cepstral coefficient extraction, feature processing and feature vector composition. Specifically, after obtaining the pre-processed training data, it can be framed to divide the pre-processed speech signal into short time frames, each frame lasting 20-40 milliseconds, so that the speech signal becomes stable in time; then windowing is performed to perform windowing operations on each frame of speech signal, and the window functions used include Hamming window, Henning window, etc., to reduce the spectrum leakage effect; then fast Fourier transform is performed on each frame of windowed speech signal to convert the time domain signal into a frequency domain signal to obtain the spectrum signal of each frame; and the Mel filter group is a group of triangular filters used to simulate the perception characteristics of the human ear to sound, through the Mel filter group By filtering the spectral signal, we can get the energy response of the Mel filter group. Then, we take the logarithm of the energy response of the Mel filter group to get the logarithmic energy spectrum, so as to compress the dynamic range of energy and enhance the details of the low-frequency part. We perform discrete cosine transform (DCT) on the logarithmic energy spectrum to get the cepstral coefficients. The discrete cosine transform can extract the envelope characteristics of the signal, and the obtained cepstral coefficients are independent of each other, which is suitable for feature representation. Finally, the cepstral coefficients are subjected to feature processing such as removing the energy attenuation coefficient and dimensionality reduction, and the processed cepstral coefficients can be combined into a feature vector as a voiceprint feature vector for the subsequent establishment of a voiceprint model.

[0112] The present application uses the Mel-frequency cepstral coefficient algorithm to extract features from the preprocessed sound data, ensuring the quality of the feature vector, providing a basis for the subsequent establishment of a voiceprint model, and improving the robustness and generalization ability of the voiceprint model. Therefore, the voiceprint model can be used to achieve non-contact fault diagnosis, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0113] In some optional implementations of this embodiment, after the step of performing diagnostic analysis according to the feature comparison result to obtain the diagnostic result of the industrial equipment to be diagnosed, the method further includes:

[0114] Receiving a data stream corresponding to the sound data in real time, and extracting incremental features;

[0115] An online learning algorithm is used to perform incremental training on the voiceprint model according to the incremental features to achieve dynamic updating of the voiceprint model.

[0116] In this embodiment, since industrial equipment will change with use and environment, the voiceprint model also needs to be regularly optimized and updated. Specifically, the data stream corresponding to the sound data of industrial equipment running in normal state and fault state can be received in real time, and processed using a message queue (such as Kafka) to ensure real-time transmission and persistent storage of data. At the same time, the speed of feature extraction can be accelerated through parallel processing, and feature extraction is performed on the data stream to obtain incremental features, and then an optimizer suitable for online learning, such as Adam or Adagr ad, is selected to incrementally train the voiceprint model using incremental features to incrementally update the model parameters, thereby realizing dynamic update of the voiceprint model to avoid frequent training of the model from scratch.

[0117] It is understandable that in order to cope with sudden large-scale data inflows, a data buffer can also be designed to prevent the system processing performance from being affected by excessive data size.

[0118] This application realizes the dynamic update of the voiceprint model through online learning, ensuring the robustness and versatility of the voiceprint model, so that when non-contact fault diagnosis is realized with the help of the voiceprint model, the accuracy of industrial diagnosis can be increased and the efficiency of industrial diagnosis can be improved.

[0119] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0120] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0121] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0122] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0123] Further references Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of an industrial diagnosis device based on voiceprint recognition, and the device embodiment is similar to Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0124] like Figure 3 As shown, the industrial diagnosis device 300 based on voiceprint recognition described in this embodiment includes: a collection module 301, an extraction module 302, a creation module 303, a comparison module 304 and an analysis module 305. Among them:

[0125] The acquisition module 301 is used to collect sound data of industrial equipment running in normal state and fault state;

[0126] An extraction module 302 is used to preprocess and extract features from the sound data to obtain a voiceprint feature vector;

[0127] Establishing module 303, used to establish a voiceprint model through a deep neural network using the voiceprint feature vector;

[0128] The comparison module 304 is used for, when receiving the target sound data of the industrial equipment to be diagnosed, performing feature comparison on the target sound data according to the voiceprint model to obtain a feature comparison result;

[0129] The analysis module 305 is used to perform diagnostic analysis based on the feature comparison result to obtain a diagnostic result of the industrial equipment to be diagnosed.

[0130] The industrial diagnostic device based on voiceprint recognition provided in the present application can automatically perform voiceprint recognition on the sound data generated by industrial equipment by establishing voiceprint models of industrial equipment in different states to achieve non-contact fault diagnosis, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0131] In some optional implementations of this embodiment, the comparison module 304 is further used to:

[0132] When the target sound data is received, preprocessing and feature extraction are performed on the target sound data to obtain a target voiceprint feature vector;

[0133] The target voiceprint feature vector is input into the voiceprint model, and the similarity between the target voiceprint feature vector and the voiceprint feature vector is calculated as the feature comparison result.

[0134] The industrial diagnostic device based on voiceprint recognition provided by the present application ensures the accuracy of feature comparison by calculating the similarity between feature vectors, and provides a data basis for subsequent diagnostic analysis, so that non-contact fault diagnosis can be achieved based on the similarity, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0135] In some optional implementations of this embodiment, the comparison module 304 is further used to:

[0136] Using metric learning technology, establishing a metric learning model according to the voiceprint feature vector;

[0137] The similarity is calculated based on the metric learning model.

[0138] The industrial diagnostic device based on voiceprint recognition provided in the present application calculates the similarity between the target voiceprint feature vector and the voiceprint feature vector stored in the voiceprint model through metric learning technology, so as to obtain more accurate feature comparison results, provide a data basis for subsequent diagnostic analysis, and enable non-contact fault diagnosis based on similarity, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0139] In some optional implementations of this embodiment, the analysis module 305 is further used to:

[0140] Comparing the similarity with a preset threshold to obtain a comparison result;

[0141] Analyze according to the comparison result to determine the device state of the industrial device to be diagnosed, wherein the device state includes a normal state and a fault state;

[0142] If the device state is a fault state, determining the corresponding fault type;

[0143] A diagnosis analysis report is generated according to the comparison result, the device status and the fault type as the diagnosis result.

[0144] The industrial diagnostic device based on voiceprint recognition provided in the present application can determine the equipment status and fault type of industrial equipment through diagnostic analysis of feature comparison results, thereby obtaining diagnostic results, realizing non-contact fault diagnosis, increasing the accuracy of industrial diagnosis, and improving the efficiency of industrial diagnosis.

[0145] In some optional implementations of this embodiment, the extraction module 302 is further used to:

[0146] De-noising, filtering and enhancing signal quality of the sound data to obtain pre-processed training data;

[0147] The training data is subjected to feature extraction according to a Mel-frequency cepstral coefficient algorithm to obtain the voiceprint feature vector.

[0148] The industrial diagnostic device based on voiceprint recognition provided by the present application ensures the quality of the feature vector by preprocessing the sound data and extracting features using the Mel-frequency cepstral coefficient algorithm, provides a basis for the subsequent establishment of a voiceprint model, and improves the robustness and generalization ability of the voiceprint model. Therefore, non-contact fault diagnosis can be achieved with the help of the voiceprint model, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0149] In some optional implementations of this embodiment, the extraction module 302 is further used to:

[0150] Performing frame division, windowing and Fourier transformation on the training data to obtain a spectrum signal;

[0151] Filtering the spectrum signal through a Mel filter bank to obtain an energy response;

[0152] Performing logarithmic operation and cepstral coefficient extraction on the energy response to obtain cepstral coefficients;

[0153] The voiceprint feature vector is formed according to the cepstral coefficients.

[0154] The industrial diagnostic device based on voiceprint recognition provided by the present application performs feature extraction on the preprocessed sound data through the Mel-frequency cepstral coefficient algorithm, thereby ensuring the quality of the feature vector, providing a basis for the subsequent establishment of the voiceprint model, and improving the robustness and generalization ability of the voiceprint model. Therefore, non-contact fault diagnosis can be achieved with the help of the voiceprint model, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0155] In some optional implementations of this embodiment, the industrial diagnosis device 300 based on voiceprint recognition is further used for:

[0156] Receiving a data stream corresponding to the sound data in real time, and extracting incremental features;

[0157] An online learning algorithm is used to perform incremental training on the voiceprint model according to the incremental features to achieve dynamic updating of the voiceprint model.

[0158] The industrial diagnostic device based on voiceprint recognition provided in the present application realizes dynamic updating of the voiceprint model through online learning, thereby ensuring the robustness and versatility of the voiceprint model, so that when non-contact fault diagnosis is realized with the help of the voiceprint model, the accuracy of industrial diagnosis can be increased and the efficiency of industrial diagnosis can be improved.

[0159] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0160] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 4 with components 41-43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital processors (DSPs), embedded devices, etc.

[0161] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0162] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk, a smart memory card (Smart Med ia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device 4. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of an industrial diagnosis method based on voiceprint recognition, etc. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.

[0163] The processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run the computer-readable instructions stored in the memory 41 or process data, such as computer-readable instructions for running the industrial diagnosis method based on voiceprint recognition.

[0164] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0165] The computer device provided in the present application can automatically perform voiceprint recognition on sound data generated by industrial equipment by establishing voiceprint models of industrial equipment in different states to achieve non-contact fault diagnosis, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0166] The present application also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the industrial diagnosis method based on voiceprint recognition as described above.

[0167] The computer-readable storage medium provided in the present application can automatically perform voiceprint recognition on sound data generated by industrial equipment by establishing voiceprint models of industrial equipment in different states to achieve non-contact fault diagnosis, thereby increasing the accuracy of industrial diagnosis and improving the efficiency of industrial diagnosis.

[0168] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0169] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. An industrial diagnosis method based on voiceprint recognition, characterized in that: The steps include: Collect sound data of industrial equipment operating in normal and fault conditions; Preprocessing and feature extracting the sound data to obtain a voiceprint feature vector; Using the voiceprint feature vector, establishing a voiceprint model through a deep neural network; When receiving target sound data of the industrial equipment to be diagnosed, performing feature comparison on the target sound data according to the voiceprint model to obtain a feature comparison result; A diagnostic analysis is performed based on the feature comparison result to obtain a diagnostic result of the industrial equipment to be diagnosed.

2. The industrial diagnosis method based on voiceprint recognition according to claim 1 is characterized in that: When receiving the target sound data of the industrial equipment to be diagnosed, the step of performing feature comparison on the target sound data according to the voiceprint model to obtain the feature comparison result specifically includes: When the target sound data is received, preprocessing and feature extraction are performed on the target sound data to obtain a target voiceprint feature vector; The target voiceprint feature vector is input into the voiceprint model, and the similarity between the target voiceprint feature vector and the voiceprint feature vector is calculated as the feature comparison result.

3. The industrial diagnosis method based on voiceprint recognition according to claim 2 is characterized in that: The step of inputting the target voiceprint feature vector into the voiceprint model and calculating the similarity between the target voiceprint feature vector and the voiceprint feature vector as the feature comparison result specifically includes: Using metric learning technology, establishing a metric learning model according to the voiceprint feature vector; The similarity is calculated based on the metric learning model.

4. The industrial diagnosis method based on voiceprint recognition according to claim 2 is characterized in that: The step of performing diagnostic analysis based on the feature comparison result to obtain the diagnostic result of the industrial equipment to be diagnosed specifically includes: Comparing the similarity with a preset threshold to obtain a comparison result; Analyze according to the comparison result to determine the device state of the industrial device to be diagnosed, wherein the device state includes a normal state and a fault state; If the device state is a fault state, determining the corresponding fault type; A diagnosis analysis report is generated according to the comparison result, the device status and the fault type as the diagnosis result.

5. The industrial diagnosis method based on voiceprint recognition according to claim 1, characterized in that: The step of preprocessing and extracting features from the sound data to obtain a voiceprint feature vector specifically includes: De-noising, filtering and enhancing signal quality of the sound data to obtain pre-processed training data; The training data is subjected to feature extraction according to a Mel-frequency cepstral coefficient algorithm to obtain the voiceprint feature vector.

6. The industrial diagnosis method based on voiceprint recognition according to claim 5 is characterized in that: The step of extracting features from the training data according to the Mel-frequency cepstral coefficient algorithm to obtain the voiceprint feature vector specifically includes: Performing frame division, windowing and Fourier transformation on the training data to obtain a spectrum signal; Filtering the spectrum signal through a Mel filter bank to obtain an energy response; Performing logarithmic operation and cepstral coefficient extraction on the energy response to obtain cepstral coefficients; The voiceprint feature vector is formed according to the cepstral coefficients.

7. The industrial diagnosis method based on voiceprint recognition according to any one of claims 1 to 6, characterized in that: After the step of performing diagnostic analysis according to the feature comparison result to obtain the diagnostic result of the industrial equipment to be diagnosed, the method further includes: Receiving a data stream corresponding to the sound data in real time, and extracting incremental features; An online learning algorithm is used to perform incremental training on the voiceprint model according to the incremental features to achieve dynamic updating of the voiceprint model.

8. An industrial diagnostic device based on voiceprint recognition, characterized in that: include: A collection module is used to collect sound data of industrial equipment operating in normal and fault conditions; An extraction module, used for preprocessing and feature extraction of the sound data to obtain a voiceprint feature vector; A building module, used to use the voiceprint feature vector to build a voiceprint model through a deep neural network; A comparison module, configured to, when receiving target sound data of the industrial equipment to be diagnosed, perform feature comparison on the target sound data according to the voiceprint model to obtain a feature comparison result; The analysis module is used to perform diagnostic analysis based on the feature comparison result to obtain a diagnostic result of the industrial equipment to be diagnosed.

9. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the industrial diagnosis method based on voiceprint recognition as described in any one of claims 1 to 7 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps of the industrial diagnosis method based on voiceprint recognition as described in any one of claims 1 to 7 are implemented.

Citation Information

Patent Citations

  • Voiceprint feature-based substation state recognition method and device

    CN112420055A

  • Power equipment fault diagnosis method, device and equipment based on voiceprint recognition

    CN113257249A

  • Voiceprint recognition method, voiceprint recognition device, voiceprint recognition equipment and storage medium

    CN116564315A

  • Power equipment fault detection method and device, electronic equipment and storage medium

    CN116778960A

  • Voiceprint recognition system for transformer

    CN117373478A

Cited By

  • Substation equipment operation state judgment method and system based on voiceprint data

    CN120126504A