429 results about "Spectrogram" patented technology

The invention, which belongs to the technical field of electromagnetic signal monitoring, relates to a rapid judgment method for a spatial spill electromagnetic signal, and the method specifically comprises the following steps: S1, receiving a spill electromagnetic signal in a space; s2, dividing the monitoring total frequency band into a plurality of sub-frequency bands with the same width, rapidly analyzing the obtained electromagnetic signal frequency spectrum in real time based on multi-scale spectrum entropy analysis, and positioning the frequency band where the characteristic signal is located; and S3, converting the frequency spectrum of the frequency band where the characteristic signal is located into a time domain signal, and simultaneously displaying a spectrogram and a time domain waveform. Through spectrum segmentation, multi-scale spectrum entropy analysis, adaptive threshold judgment and frequency domain-time domain combined display, the problems of poor real-time performance, weak anti-noise capability and single dimension in the prior art are solved.

The present invention relates to the technical field of fault diagnosis, and provides a fault diagnosis method for a wind-turbine drive train based on time-frequency plane expectation maximization. The method comprises the following steps: configuring a cylindrical MEMS acceleration sensor on a drive chain of a wind turbine, and acquiring a vibration signal of the drive chain of the wind turbine via the cylindrical MEMS acceleration sensor; adopting a spectrogram-zero-based unsupervised classification method, acquiring a vibration signal and a rotational speed signal of the wind turbine during operation, generating a time-frequency representation of the vibration signal by using a short-time Fourier transform, and extracting spectrogram zeros and performing unsupervised classification to implement denoising processing of the vibration signal; on the basis of a multi-component signal estimation method under time-frequency plane expectation maximization, accurately estimating an instantaneous frequency and an instantaneous amplitude of a multi-component signal on a time-domain plane of the denoised signal; and performing order spectrum analysis to identify vibration signal characteristics of the drive chain of the wind turbine, so as to achieve fault diagnosis of the drive chain of the wind turbine under variable rotational speeds.

The invention discloses a wind turbine generator impeller anomaly detection method and system based on sound vibration signal identification. According to the method, firstly, impeller response is collected through a microphone / vibration sensor, and a Mel spectrogram is generated; then, a Teager-Kaiser energy operator and self-correlation analysis are utilized, and under the condition of not depending on a rotating speed signal, the rotating period of the impeller is recognized, and frequency spectrum segmentation is carried out; calculating a cross-period Mel frequency spectrum dynamic deviation, and normalizing the cross-period Mel frequency spectrum dynamic deviation through an amplitude correction coefficient related to the rotating speed; generating a periodic coherent energy diagram by adopting an improved structural similarity algorithm; performing filtering enhancement by using a harmonic resonance template, performing morphological deconstruction and parameterization on an abnormal region in the graph, and extracting geometric features; and finally, calculating a comprehensive abnormal score based on the multi-dimensional features and realizing automatic early warning. According to the method, the problems of variable working condition interference and rotating speed dependence are effectively solved, and accurate and stable detection of early abnormality of the impeller can be realized.

The invention discloses a robust radio frequency fingerprint identification method based on Barlow Twins domain adaptation. Aiming at the problems of training and deployment environment distribution offset and cross-domain identification performance reduction caused by wireless channel multipath and time-varying characteristics in the prior art, the invention provides a domain adaptation framework combined with feature decoupling. The method comprises the following steps: firstly, converting a received time domain signal into a short-time Fourier transform spectrogram to reserve a time-frequency structure; then, a double-end convolutional encoder is constructed, bottom layer features are extracted through a shared backbone network, and device fingerprints and channel features are obtained in an identity branch and a channel branch respectively; in order to realize thorough decoupling of the two types of features, Barlow Twins-based independence regularization is introduced, and the identity features and the channel features are statistically orthogonal by minimizing a cross-correlation matrix of the identity features and the channel features, so that purer fingerprint features are obtained, channel interference is effectively eliminated, and generalization and recognition precision of the model under an unknown channel are remarkably improved.

The invention discloses a transformer abnormity identification method based on voiceprint feature analysis, and belongs to the field of power equipment state monitoring and intelligent diagnosis. The method comprises the following steps: firstly, analyzing an iron core acoustic mechanism based on a magnetostrictive effect, and establishing a three-dimensional model through finite element simulation to obtain vibration and sound field characteristics; in a complex substation environment, a hybrid noise reduction method combining density peak clustering and a CEEMDAN-wavelet threshold is provided, and the signal-to-noise ratio is effectively improved. Then extracting Mel-frequency cepstrum coefficients (MFCC) and spectrum features, and performing local linear embedding (LLE) dimension reduction to form a compact feature set; in the recognition stage, a convolutional neural network framework is designed, specifically, a spectrogram and an energy spectrum are modeled through a two-dimensional CNN, an MFCC tensor obtained after dimensionality reduction is modeled through a three-dimensional CNN, and accurate diagnosis of mechanical faults such as core looseness is achieved. The method has the advantages of being non-contact, anti-noise and high in recognition precision, and real-time diagnosis and early warning of mechanical abnormity of the transformer can be achieved under complex working conditions.

The invention discloses a signal reconnaissance, positioning and interference method based on deep learning, and belongs to the technical field of electronic countermeasures. The method comprises the following steps: deploying distributed reconnaissance nodes to obtain broadband electromagnetic signals; generating a spectrogram through time-frequency transformation, and identifying a fixed frequency / frequency hopping signal based on a deep learning model; fusing multi-node reconnaissance results to realize target situation awareness; the optimal positioning cluster is selected through dynamic networking, and the target position is calculated in combination with the arrival angle and time difference information; generating an optimal interference strategy based on the target situation and the position trajectory; networking cooperative interference is implemented; behavior parameters are collected in real time, the interference effect is evaluated through a machine learning model, and the strategy is dynamically adjusted. According to the method, the technical effects of high-efficiency and self-adaptive signal reconnaissance, positioning and interference in a complex electromagnetic environment are achieved.

The invention relates to the technical field of power equipment state monitoring, and discloses a power equipment fault detection method and device, equipment and a medium, and the method comprises the steps: obtaining a voiceprint signal of power equipment, carrying out the time-frequency transformation to obtain an original logarithmic Mel spectrogram, inputting a multi-scale context sensing auto-encoder model, and outputting a reconstructed spectrogram. According to the model, multi-scale long-range dependence features of voiceprints in time and frequency dimensions are respectively extracted by using a double-flow expansion convolutional network, and complete spectrum reconstruction is carried out based on the extracted features; and calculating an abnormal score based on a reconstruction difference degree between the original logarithmic Mel spectrogram and the reconstructed spectrogram, and when the abnormal score exceeds a dynamic threshold value, judging that the equipment has a fault. Compared with the prior art, the problems that weak fault features are difficult to extract and reconstruction details are fuzzy under strong background noise are solved, and high-robustness non-contact fault detection is achieved.

The invention discloses a multi-mode control method for voice interaction and gesture recognition of an atmosphere lamp, and belongs to the field of intelligent control. The method comprises the following steps: S1, constructing a multi-mode sensing module, and arranging a bidirectional linear array microphone, a dual-lens infrared depth camera and a space calibration laser transmitter; s2, extracting voice features by adopting an extended spectrogram domain attention network, and generating a first semantic vector; s3, dynamic attitude features are extracted through the time-airspace sparse convolutional network; s4, performing modal alignment and cooperative coding on the two semantic vectors based on a bidirectional gating fusion mechanism, generating a joint interaction instruction vector, and solving modal conflicts by a time sequence synchronization and confidence regulation strategy; s5, inputting the vector into a multi-task intention classifier, analyzing a user intention and outputting a unique atmosphere lamp control code; and S6, transmitting the control code to a driving unit to realize light control. The beneficial effects are that voice and gesture fusion control is realized, and atmosphere lamp interaction intelligence and response precision are improved.

The invention discloses an electrolytic tank performance analysis method based on polarization feature recognition. The method comprises the steps that S1, electrolytic tank parameters are set, electrochemical testing is conducted, and EIS impedance data are obtained; s2, preprocessing the EIS impedance data: performing Kramers-Kronig verification on the EIS impedance data, calculating a relative residual error of the data, and selecting a regularization parameter lambda; s3, applying the EIS impedance data to DRT quantitative analysis: converting the EIS impedance data into a DRT spectrogram based on discretization of a radial basis function and a Tikhonov regularization method; and S4, performing DRT spectrogram identification by using a DRT method, establishing an ECM model, and evaluating the influence of metal cation pollution on PEMWE performance degradation by using the ECM model. According to the electrolytic cell performance analysis method based on polarization feature recognition, the core problem that all polarization processes and the evolution rule of the polarization processes in performance degradation are difficult to accurately analyze under the condition of high current density in the prior art is solved.

The invention relates to the field of radio interference intelligent identification and positioning, in particular to a multi-source data fusion water radio interference identification and positioning method and system, and the method comprises the following steps: collecting multi-source data; performing multi-source data fusion processing; constructing a rule base, and screening out suspected abnormal signals from the frequency spectrum situation map through threshold judgment and rule matching; a support vector machine model is called, multi-dimensional static signal features are used as input, a classification hyperplane is constructed through a radial basis kernel function, and normal signals and abnormal signals are distinguished from suspected abnormal signals; calling a decision tree model, and performing scene judgment on the suspected abnormal signal to identify the abnormal signal; calling a convolutional neural network model, converting the time domain signal into a spectrogram through short-time Fourier transform, and extracting texture features through multilayer convolution to identify an abnormal signal; through the method and the system, real-time identification and high-precision positioning of interference signals can be realized.

A radio frequency fingerprinting method and system based on a convolution-attention mechanism and multi-packet inference are disclosed, and belong to the technical field of communication networks and artificial intelligence. The radio frequency fingerprinting method based on the convolution-attention mechanism and multi-packet inference includes: step 1: capturing a device transmission signal, and preprocessing the transmission signal to obtain a spectrogram; step 2: constructing a radio frequency fingerprinting model, and inputting the obtained spectrogram into the radio frequency fingerprinting model for training, to obtain a trained radio frequency fingerprinting model; and step 3: performing prediction by employing the trained radio frequency fingerprinting model, to obtain a final prediction result.

The invention provides a bone conduction speech enhancement method based on a double-flow self-attention fusion network. The bone conduction speech enhancement method comprises the following steps: acquiring and preprocessing a bone conduction speech signal to be processed; a double-flow self-attention fusion network is constructed, the double-flow self-attention fusion network comprises a double-flow encoder, a cross-modal feature fusion module, a multi-scale decoder and a waveform reconstruction module which are connected in sequence, the double-flow self-attention fusion network is trained, a trained model is obtained, and the performance of the model under different conditions is evaluated. According to the method, a double-flow structure is used for processing time domain information and frequency domain information respectively, a self-attention mechanism is introduced, so that the model can capture a long-distance dependency relationship in each mode, and the feature quality is improved. According to the invention, the time-frequency information can be better integrated through the cross-modal feature fusion module. According to the invention, the multi-scale loss function supervises the output of different levels of the decoder, so that the network can learn effective spectrogram representation at different resolutions, and a finer spectrum structure can be recovered.

A computer-implemented method and an apparatus for automatic modulation recognition that enables the detection and identification of modulation schemes in received raw signals with a signal receiving unit (70) without prior information about the raw signal detail, characterized by; comprising at least one computing unit (10) configured to transforming received raw signals from time domain to frequency domain including the noise in the signal with segmenting the signal and computing its modulation in multiple image with the spectrogram extraction process in order to capture temporal dependencies and sequential information by treating the raw signals received from signal receiving unit (70) as images; augmentation of the data for increasing the dataset size for increasing the accuracy with the limited data; training the data for enabling the network to learn spatiotemporal relationships based on a predefined or a precalculated signal-to-noise ratio (SNR) level; applying the data to one algorithm of two, which are convolutional neural network (CNN) and convolutional neural network (CNN) long short-term memory network (LSTM) hybrid algorithm.

The invention discloses a time-frequency fusion power line noise classification modeling method, system and device based on deep learning, and a medium, and belongs to the technical field of power line noise classification modeling, and the method comprises the steps: constructing a data set containing multiple types of typical power line noise signals; performing short-time Fourier transform on the noise signal to obtain a spectrogram, and extracting local texture features in the spectrogram; extracting global time sequence characteristics of the original time domain waveform of the noise signal; splicing the local texture features and the global time sequence features to form a fusion feature vector, inputting the fusion feature vector into a full-connection neural network, and outputting a multi-label noise prediction result; and training the full-connection neural network, and updating network parameters by optimizing a binary cross entropy loss function. According to the method, different characteristics and generation mechanisms of each type of noise are considered, through fusion of time domain information and frequency domain information, the characteristics can be extracted more comprehensively, the space relation and mode of the time domain and the frequency domain are effectively captured, and the adaptability and accuracy of the model to a noise classification task are improved.

The invention discloses a bimodal identity authentication method based on voice and ultrasonic signals. The method comprises the following steps: transmitting an orthogonal frequency division multiplexing modulated ultrasonic wave by using a built-in loudspeaker of equipment, and synchronously acquiring a mixed signal containing voice and ultrasonic reflection; after separation, converting the speech signal into a speech spectrogram, and carrying out vocal organ positioning, segmented de-trending and multi-scale spectrogram conversion on the ultrasonic signal to extract anti-interference pronunciation action features; the method comprises the following steps: respectively extracting depth features of voice and ultrasound through a double-flow neural network, and carrying out dynamic weighted fusion according to signal quality by adopting a quality sensing channel gating fusion module to generate an identity embedding vector irrelevant to speaking content; in the registration stage, a speaker template is constructed, and a personalized threshold value is calculated; and in the authentication stage, identity judgment is realized through cosine similarity comparison. According to the method, the robustness of the system in a noise environment and the defense capability of the system on spoofing attacks such as playback and synthesis are remarkably improved by fusing the voice content and the ultrasonic living body action characteristics.

The invention relates to a multi-source partial discharge fault spectrogram detection and separation method and device and a medium. The method comprises the steps that a multi-source partial discharge fault spectrogram is collected and preprocessed; inputting the preprocessed multi-source partial discharge fault spectrogram into a pre-trained detection separation model to obtain a separated partial discharge fault spectrogram; wherein the detection separation model is pre-trained based on a multi-source partial discharge fault spectrogram data set, the multi-source partial discharge fault spectrogram data set is subjected to sample expansion through a Wasserstein generative adversarial network with gradient penalty, and the detection separation model uses a VoVNet network as a backbone feature extraction network. By generating a multi-scale feature map corresponding to different sizes of fault features in multi-source partial discharge, the detection and separation model effectively detects different types and sizes of fault features in multi-source partial discharge, and a multi-source partial discharge fault spectrogram is separated. Compared with the prior art, the characterization and detection capability of the model on the partial discharge characteristics of the small target is remarkably enhanced.

The invention relates to the technical field of voice authentic identification, can be applied to the fields of finance and medical treatment, and discloses a voice authentic identification method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring audio data, and processing the audio data to obtain audio spectrogram data and audio time domain features; performing multi-layer two-dimensional convolution processing on the audio frequency spectrogram data to obtain convolution feature map data; determining a plurality of feature units according to the convolution feature graph data; performing aggregation processing on the plurality of feature units to obtain a frequency domain embedding representation; performing cross attention calculation on the audio time domain features according to the frequency domain embedded representation to obtain enhanced time domain features; performing cross attention calculation on the frequency domain embedded representation according to the audio time domain feature to obtain an enhanced frequency domain feature; performing fusion processing on the enhanced time domain feature and the enhanced frequency domain feature to obtain a fusion feature; and discriminating the fusion features, and determining to obtain an authentic identification result. The method has the advantages of accurate result, good robustness, strong generalization ability and the like.

A system (100) for anomaly detection in an industrial drivetrain (104) comprising various mechanical elements driven by a three-phase electrical motor (102). The system comprises a sensor system (101) for obtaining electrical current and voltage waveform data by measuring electrical currents and voltages at the three-phase motor, and a computing system. The electrical waveform data contains frequency components relating to vibrations of mechanical elements of the drivetrain. The computing system performs the following steps: processing the electrical waveform data to obtain related frequency spectra, thereby enhancing traceability of the frequency components; combining the frequency spectra into a combined frequency spectrum, thereby further enhancing traceability of the frequency components; forming a spectrogram by monitoring the combined frequency spectrum over time; and tracing the frequency components in the spectrogram for anomaly detection of the related mechanical elements of the industrial drivetrain.

The invention discloses a double-channel voice stream real-time identification method, system and application, and the method comprises the steps: carrying out the preprocessing of collected VOIP call double-channel audio, and maintaining the time sequence synchronization; extracting Mel-frequency cepstral coefficient features and speech spectrogram features of the preprocessed audio, and inputting the spliced features into a Transform deep neural network model for stream speech recognition to obtain a two-way character sequence; generating a unique identifier based on channel identification and voice energy difference, and establishing a corresponding relation with the character sequence; carrying out punctuation prediction and text standardization by utilizing an LSTM-based model, sorting and aligning character sequences according to timestamp fields, and generating a time sequence dialogue stream; and performing anomaly detection and / or storage management on the time sequence dialogue stream to realize real-time quality inspection and agent assistance. According to the invention, synchronous recognition and role distinguishing of double-channel voice are realized, the recognition delay is low, and the recognition accuracy, the detection precision and the real-time performance are high; and high-efficiency management and safety compliance of data are realized by combining distributed encryption storage.

The invention discloses a cable force inversion method based on vision measurement and Hilbert-Huang transformation, and relates to the technical field of structural vibration measurement. The method comprises the steps of system deployment, video acquisition and preprocessing, sub-pixel displacement extraction, EEMD signal decomposition, Hilbert time-frequency spectrum construction, improved string vibration formula inversion cable force and health degree evaluation. The method supports abnormal grading processing and long-term structure health monitoring, has engineering practicability, and is suitable for cable force dynamic monitoring and safety evaluation of structures such as bridges and cableways.

The invention relates to the field of seismology, signal processing and artificial intelligence, namely to automated methods and systems for early detection of earthquakes based on the analysis of seismic data, and can be used to recognize and classify longitudinal waves (P-waves) preceding the main seismic shocks, using deep learning and computer vision methods. The aim of the present invention is to create an automated early earthquake detection system using the classification of seismic signal spectrograms generated by the complex Morlet wave CWT using the YOLO deep neural network architecture. The technical result is an increase in the accuracy and speed of early earthquake detection by analyzing the time-frequency characteristics of seismic signals and automatically localizing P-wave signatures using a neural network model. The device includes a seismic sensor, an analog-to-digital converter, and a microprocessor implementing time-frequency analysis and neural network detection algorithms. The seismic signal is recorded in real time, digitized, segmented into time intervals, and subjected to preliminary digital processing, including noise filtering and amplitude normalization. Each time interval is converted into a time-frequency representation using a continuous wavelet transform, generating a two-dimensional distribution of signal energy over time and frequency. Based on the obtained data, a spectrogram is generated and fed to the YOLO neural network detection model, which is capable of automatically detecting and localizing longitudinal P-wave signatures. Based on the neural network analysis, a determination is made regarding the presence of a P-wave and its arrival time is determined. If a predetermined threshold is exceeded, an early warning signal is generated. The system provides for data accumulation and the possibility of subsequent retraining of the neural network model.

Systems and techniques described herein relate to a diffusion-based model for generating converted speech from source speech based on target speech. For example, a device may extract first rhythm data from input data, and may generate content embedding based on the input data. The device may extract second rhythm data from the target speech, generate a speaker insert from the target speech, and generate a rhythm insert from the second rhythm data. The device may generate converted rhythm data based on the first rhythm data and the rhythm embedding. The device may then generate a converted spectrogram based on the converted rhythm data, the speaker embedding, and the content embedding.

The invention discloses a frequency modulated continuous wave cable local insulation defect detection method, which comprises the following steps: S1, realizing an FMCW method by using an FMCW cable fault positioning system, performing spectrogram function extraction and analysis on an incident signal and a reflected signal at a certain moment in the FMCW method to obtain a frequency difference between the incident signal and the reflected signal at the certain moment, and determining the frequency difference between the incident signal and the reflected signal at the certain moment; therefore, the position information from the cable defect to the head end of the cable is obtained, and effective fault diagnosis is realized. S2, time domain, frequency domain and power spectrum characteristic index extraction is carried out on the spectrogram function analysis signals, and index data of the analyzed signals are known more comprehensively; the method solves the problems that in the prior art, a reflection method is limited by frequency domain reflection signal amplitude attenuation, the detection distance is short, the requirement for long-distance cable detection cannot be met, FFT positioning spectrum covering information is single, interference of environmental factors is likely to happen, and cable state evaluation is not comprehensive enough.

The invention discloses an audio-visual joint detection and classification method for a faulty carrier roller of a belt conveyor, which belongs to the field of intelligent monitoring and fault diagnosis of the belt conveyor, and comprises the following steps: synchronously acquiring an image signal and an audio signal of the belt conveyor, and carrying out time sequence alignment processing to obtain synchronous image frames and audio clips; processing the audio clip by adopting a time-frequency transformation method to obtain a two-dimensional audio spectrogram; performing channel splicing on the image frames and the corresponding two-dimensional audio spectrograms, and inputting the image frames and the corresponding two-dimensional audio spectrograms into a single feature extraction network to obtain cross-modal fusion features; based on the cross-modal fusion features, pixel-level positioning and fault type identification of the fault carrier roller are carried out through a double-branch structure, and a segmentation mask and a fault type are obtained; and performing end-to-end joint training on the feature extraction network and the double-branch structure by adopting a multi-task joint loss function based on the label so as to synchronously optimize the positioning and classification tasks. According to the invention, the positioning precision of the fault sounding carrier roller is obviously improved.

The invention discloses an intelligent traffic flow prediction method and device based on spatio-temporal feature information fusion, and belongs to the technical field of intelligent traffic prediction.The method comprises the steps that obtained historical traffic flow data are input into a pre-constructed DSTMGCN model; processing the input traffic flow data through a time attention layer in the DSTMGCN model, and extracting time feature information; inputting the time feature information and historical traffic flow data into a spatial multi-graph convolution layer; based on a spectrogram convolution method, performing convolution operation on the time feature information and the historical traffic flow data by using a predefined distance map, an adjacent map and an association map respectively so as to extract features fused with a space-time dependency relationship; and a final traffic flow prediction value is generated. According to the method, the distance graph, the adjacency graph and the association graph are introduced, so that potential relevance among different nodes is effectively mined, and the defects of a traditional method in spatial information extraction are overcome.

The application provides a continuous human action recognition method based on point cloud data and range Doppler spectrograms, comprising the following steps: collecting echo containing multiple continuous human actions to obtain sampled intermediate frequency signals, and obtaining corresponding range Doppler spectrograms and point cloud data after processing; building a PVT-BiLSTM network, wherein the PVT-BiLSTM network comprises a Pointnet-BiLSTM module, a ViT module and a fusion module; dividing the range Doppler spectrograms and the point cloud data into a training set and a verification set, and inputting them into the PVT-BiLSTM network for training and verification to obtain a trained PVT-BiLSTM network model, and then using the trained PVT-BiLSTM network model to recognize continuous human actions. The method can significantly improve the recognition performance of continuous human actions.

This invention provides a quantum-derived Newton-Raphson optimal fractional-order spectrogram generation method, comprising the following steps: Step 1: Acquire the original audio signal and construct a fractional-order spectrogram based on fractional Fourier transform (FRFT); Step 2: Perform nonlinear scaling compression on the fractional-order spectrogram using a Mel filter bank to generate a fractional-order Mel spectrogram; Step 3: Construct an adaptive optimization framework with information entropy minimization as the objective function to measure the information fidelity between the spectrogram and the original signal; Step 4: Use the quantum-derived Newton-Raphson optimization algorithm (QNRBO) to globally optimize the fractional-order order, frame length, and frame shift hyperparameters to generate the optimal fractional-order spectrogram; Step 5: Input the optimal fractional-order spectrogram into a downstream speech recognition model. This technical solution aims to systematically solve core problems such as insufficient traditional time-frequency representation capabilities, rigid hyperparameter configuration, limited optimization algorithm performance, and feature-task disconnect.

The invention discloses a circuit reduction method and system based on sparse processing, and the method comprises the steps: A, carrying out the node elimination of a circuit matrix through employing a time constant balance reduction (TICER) algorithm, and obtaining a matrix which is smaller in scale and is dense due to filling; b, deleting redundant non-zero elements by using a spectrogram sparsification algorithm, and meanwhile, keeping the core characteristics of the matrix to obtain a sparsification matrix; and C, judging the density of the matrix, if the sparse effect reaches the expectation, outputting a new matrix after reduction, otherwise, returning to the step of the initial matrix. By the adoption of the method, circuit reduction can be carried out through the method of combining spectrogram sparsification and the TICER algorithm, the matrix density degree is effectively reduced, and the simulation speed is greatly increased; and the circuit reduction effect can be ensured by setting a rollback mechanism.

The invention provides a self-adaptive wavelet noise reduction method and system based on mass spectrum signal processing, and the method comprises the following steps: S1, collecting a data set outputted by a mass spectrometer, and inputting the data set into a noise reduction process in the form of a text document; s2, reading related data in the text document, analyzing spectral peak characteristic parameters, and initializing related parameters; s3, performing multilayer wavelet decomposition on the document data, and outputting an approximate component and a detail component of each layer; s4, calculating the noise intensity of different layers according to the output approximate component and the detail component, and adaptively calculating a wavelet threshold value based on the signal length and the noise intensity; s5, threshold processing is applied to the detail components, and wavelet signals are reconstructed; and S6, calculating signal-to-noise ratios under different decomposition layer numbers, and selecting the optimal decomposition layer number to output the mass spectrum data after noise reduction. According to the method, the related spectrogram information output by the mass spectrometer is denoised, and the signal-to-noise ratio of the mass spectrum data is improved while the spectrum peak information is reserved to a great extent.

This application provides a method and system for adding and extracting hidden watermarks in audio, including: obtaining a first waveform file corresponding to the audio to be synthesized; generating user data corresponding to the first waveform file; performing frequency domain transformation on the first waveform file to obtain a first Mel spectrogram; adding the user data to the first Mel spectrogram to obtain a second Mel spectrogram containing the hidden watermark; and performing restoration processing on the second Mel spectrogram to obtain a second waveform file containing the hidden watermark. By injecting a hidden watermark into the Mel spectrogram of the synthesized audio, the original user data can be obtained through watermark extraction. This facilitates tracing the source based on the user data, understanding information such as the data producer or user, enhancing the monitoring and management capabilities of disseminated audio content, and providing effective copyright protection for audio content. This avoids the illegal application of synthesized audio content and reduces the harm caused to the speaker by synthesized audio content.