196 results about "Acoustic spectrum" patented technology

The invention relates to the technical field of automatic welding control, and particularly discloses a multi-station automatic welding cooperative control method for electric iron accessories, which realizes full-closed-loop accurate control of a welding process through multi-sensor data fusion and an intelligent optimization algorithm. According to the method, an infrared thermal imager, an acoustic emission sensor and an ultrasonic thickness measuring device are adopted to collect molten pool temperature field, sound wave frequency spectrum and fusion depth data in real time, after feature extraction, the data are input into a multi-target optimization model, and a Pareto optimal solution set is generated with the fusion depth, the width of a heat affected zone and phase change latent heat as optimization targets; dynamically adjusting a segmented welding-back path and process parameters on the basis of an optimization result, and triggering model iterative correction through ultrasonic online monitoring; the central control unit adopts space voxelization monitoring and a three-level collision avoidance strategy to coordinate operation of the multiple mechanical arms, and collision emergency protection is achieved in combination with flexible pressure sensing.

The invention relates to a multi-agent cooperation strategy generation method, device and equipment and a medium, and the method comprises the steps: separating acoustic spectrum features and text semantic features of conference voice through environment perception processing, solving a cross-modal information conflict problem, and generating an accurate semantic understanding result; identifying the essence of the problem based on task analysis, associating the responsibility field, and constructing a classifiable problem point set; calling an agent capability library to dynamically match problem requirements, and generating a candidate agent list; quantifying a problem influence range and a decision time limit through weighted emergency scores, and generating a priority-sorted agent sequence; screening and confirming a core problem point and a primary agent; and finally, generating an executable cooperation scheme through multi-agent collaborative optimization. According to the method, the problems of incomplete feature extraction, task allocation delay and resource conflict in the prior art are solved, and the operability and decision-making efficiency of a cooperation strategy are remarkably improved.

The invention discloses an unmanned aerial vehicle positioning method based on a Bayesian network, and the method comprises the steps: collecting unmanned aerial vehicle data, carrying out the preprocessing, obtaining sound wave data, motion data, infrared image data and environment data, carrying out the spectrum analysis based on the sound wave data, obtaining sound spectrum time sequence data, and carrying out the positioning of the unmanned aerial vehicle. Constructing a probability density thermodynamic diagram and extracting a first range through a sparse variational Gaussian process according to time, frequency and energy three-dimensional information of the sound spectrum time sequence data and environment data, and obtaining first positioning data through a Bayesian network according to the first range and motion data, and performing adaptive wavelet down-sampling on the infrared image data and the sound spectrum time sequence data in the lateral direction to obtain lateral positioning data, and performing deviation correction on the first positioning data by using the lateral positioning data to obtain second positioning. According to the method, through space-time synchronization, adaptive processing and probabilistic reasoning of sound spectrum time sequence data, the positioning precision and stability in a complex environment are improved.

The invention discloses a pipeline defect identification method and device, a storage medium and equipment, and belongs to the technical field of pipeline defect detection. The method comprises the following steps: S1, acquiring a time sequence sound signal acquired by an acoustic sensor arranged on the inner wall and / or the outer wall of a pipeline; s2, performing spectrogram extraction on the time sequence sound signal to obtain a Mel spectrogram and a LOFAR spectrogram; and S3, performing feature extraction on the Mel spectrogram, the LOFAR spectrogram and the time sequence sound signal through a deep learning model of multi-modal signal fusion to obtain a feature vector containing spectrogram features and time sequence features, and performing classification according to the feature vector to obtain a classification result of pipeline defects. According to the method, the acoustic signals are converted into different types of spectrograms, the original time sequence signals are combined, feature extraction, fusion and classification are carried out through a deep learning model, complementary information of the spectrograms and the time sequence signals is fully utilized, the limitation of a single recognition mode is overcome, and the accuracy and robustness of pipeline defect detection are improved.

The invention discloses a spectrogram feature enhancement method for arc fault detection and a series AC arc fault detection method and system, and belongs to the technical field of arc fault detection. The invention aims to solve the problem that fault characteristic information carried by current is easily covered by a so-called shielding load in the existing arc fault detection. The method comprises the following steps: performing short-time Fourier transform on time sequence data of a current signal under a load condition to obtain time-frequency information Ct, f of the current signal, and visualizing the time-frequency information Ct, f to obtain a spectrogram of an original current signal; based on the spectrogram of the original current signal, the spectrogram is processed according to St, f = 10 * log10 (Ct, f), an enhanced spectrogram is obtained, and the enhanced spectrogram is used for arc fault detection. In the detection process, an arc fault detection model constructed based on the deep learning technology is adopted to carry out fault detection, and the arc fault detection model outputs an arc fault detection result.

The invention discloses an automatic processing device and method for PCB lamination, and the method comprises the steps: depositing an erodable film on a lamination material layer and arranging a cavity concentric with a through hole array before the lamination of a PCB, and collecting acoustic resonance frequency data to obtain a reference matrix; dynamic dot frequency tracking processing is carried out on the multi-point acoustic time sequence signals in the pressing heating stage, and a through hole instantaneous acoustic track library is obtained; performing coupling judgment on time-temperature-pressure ternary working condition data according to the trajectory library to obtain a blockage risk matrix; carrying out time delay and gradient adjustment on a pressing temperature and pressure curve based on the risk matrix; and carrying out three-dimensional inversion processing on the whole board broadband acoustic spectrum collected before cooling, calculating through hole filling rate distribution and judging the quality of the board. According to the scheme, the resin filling state of the micro through hole can be monitored in real time in the PCB pressing process, an acoustic detection channel is constructed through the erodable film and the cavity structure, early warning and active intervention of the blocking risk are achieved, and the pressing quality is ensured.

The invention relates to the technical field of oil and gas exploitation, and discloses an oil and gas well water production position positioning method, device and equipment based on distributed optical fiber sound waves and a medium, and acoustic signals corresponding to a plurality of production profiles can be collected through a distributed acoustic sensor arranged in an oil and gas shaft. And performing preprocessing, time-frequency conversion and normalization on the acoustic signal corresponding to each production profile to obtain a normalized acoustic spectrum corresponding to each production profile. And extracting a characteristic wave crest section from the normalized acoustic spectrum corresponding to each production profile. And carrying out reverse normalization on the characteristic wave crest segment corresponding to each production profile to obtain a reverse normalization wave crest segment. And based on the water phase response frequency band, determining a matched first wave crest section in the anti-normalized wave crest section corresponding to each production profile, and determining the production profile corresponding to each first wave crest section as a water production position. According to the invention, intelligent detection of the water production position of the oil and gas well can be realized, and the detection precision of the water production position is effectively improved.

The invention relates to the technical field of arc fault diagnosis, and provides an energy storage system arc fault diagnosis method and system based on voiceprint detection, and the method comprises the following steps: collecting a voiceprint signal in real time, and carrying out the anti-interference preprocessing, and obtaining a time-frequency domain spectrogram; identifying the time-frequency domain spectrogram based on a multi-branch convolutional neural network and weighting the time-frequency domain spectrogram to obtain a time sequence feature sequence; inputting the time sequence feature sequence into a time sequence convolutional network model, and outputting global context features; performing dynamic weighted pooling on the global context features to obtain key frame feature vectors representing the arc fault; and inputting the key frame feature vector into an arc fault diagnosis classification model to obtain an arc fault type, an arc fault severity grade and sound source positioning information, and generating a corresponding alarm signal and a control instruction based on the arc fault severity grade. According to the invention, the detection sensitivity and the anti-interference capability are improved, and the safety protection level and the fault early warning capability of the energy storage system are improved.

The invention relates to the technical field of gas sensing detection, and discloses a gas detection method and system based on a quartz tuning fork enhanced photoacoustic spectrum. The method comprises the following steps: constructing a standardized continuous tuning fork signal data stream; dynamically segmenting the data stream based on the gas concentration interval, and extracting time domain and frequency domain features in parallel in each segment to form an initial feature vector; dynamically adjusting an anomaly judgment threshold according to the jump amplitude of the adjacent segment feature vectors, and scanning and screening out an effective analysis interval of a stable concentration state according to the anomaly judgment threshold; performing cross-domain coupling compensation on the feature vectors in the effective interval to generate enhanced feature vectors; and finally, performing reverse traceability reasoning through a gas leakage propagation model based on the enhanced feature vector, and outputting a leakage source position identifier. The device can adapt to a dynamically changing concentration field, and the reliability and traceability positioning capability of gas detection in a complex environment are improved.

The invention discloses a dynamic encryption transmission method and system for sound-light alarm signals, and relates to the field of data encryption, and the method comprises the steps: obtaining a dynamic environment entropy value; generating a pseudo-random chaotic sequence; obtaining a helium-oxygen mixed gas density parameter in the dynamic environment entropy value; segmenting a preset standard alarm waveform into a plurality of signal waveforms in a time domain, and generating a distorted signal waveform according to the helium-oxygen mixed gas density parameter; performing encryption modulation on the distorted signal waveform to generate an encrypted distorted signal waveform; determining a sending sequence and a sending interval of the encrypted distortion signal waveform by combining a pseudo-random chaos sequence based on an impedance fluctuation value and background acoustic spectrum characteristics in the dynamic environment entropy; generating a dynamic modulation signal based on the transmission sequence and the transmission interval; the dynamic modulation signal is sent to a water surface console through an umbilical cable; and receiving the guide pulse, performing phase conjugation processing on the guide pulse, generating an echo verification signal, and transmitting the echo verification signal back to the water surface console.

The invention discloses an automatic charging processing method and device for a charging robot, and relates to the technical field of robot energy management, and the method comprises the following steps: constructing an acoustic array, collecting high-frequency background sound wave signals in a working area of the charging robot in real time, and extracting sound spectrum energy distribution; and identifying an interference frequency band coinciding with the working frequency of the ultrasonic sensor, and calculating the intensity level of the interference frequency band according to the sound spectrum energy intensity and the duration time in the interference frequency band. According to the method, a closed-loop system of interference perception-risk assessment-control response is constructed, and dynamic adjustment of perception confidence and output of blindness risk indexes are realized through multi-modal identification consistency analysis and historical deviation tracking. Once the threshold value is exceeded, the system executes frequency hopping and strategy optimization, frequency band shielding is effectively avoided, the sensing stability, path accuracy and operation safety of the charging robot in the sound interference environment are improved, and good application prospects are achieved.

The invention belongs to the technical field of geological analysis, and discloses a ground subsidence analysis method and device, and the method comprises the steps: matching a sensor layout strategy corresponding to a foundation subsidence risk and meteorological data of a monitoring region; acquiring land subsidence data of the monitoring area according to a sensor; performing time-frequency analysis on the land subsidence data to obtain a geoacoustic spectrogram corresponding to the land subsidence data; inputting the geoacoustic spectrogram into a preset neural network, and outputting a settlement semantic vector corresponding to the geoacoustic spectrogram; acquiring and analyzing historical land subsidence data, and constructing a subsidence rule corresponding to the historical land subsidence data; and inputting the settlement semantic vector and the settlement rule into a preset space-time analysis model, and predicting to obtain a future settlement trend of the monitored area. By using the method disclosed by the invention, the problem that all risks in a certain area cannot be checked when geological data is monitored based on local monitoring points can be solved, and the problem that the geological risk analysis efficiency is relatively low can be solved.

The invention relates to the field of photovoltaic equipment abnormity early warning, and provides a photovoltaic equipment temperature abnormity prediction method and device based on a digital twinborn model, and the method comprises the steps: obtaining an environment acoustic signal, meteorological data, and the actually measured temperature data of photovoltaic equipment; performing acoustic spectrum feature extraction according to the environmental acoustic signal to obtain a virtual environmental parameter set; sending the meteorological data and the virtual environment parameter set to a preset digital twinborn model for processing to obtain an initial temperature predicted value set; determining a temperature prediction value distribution diagram according to the initial temperature prediction value set and actually measured temperature data of the photovoltaic equipment, and comparing a temperature prediction result with an abnormal temperature threshold value to obtain a temperature abnormal risk point; and processing according to the temperature prediction value distribution diagram, the temperature anomaly risk points and the meteorological data to obtain temperature anomaly early warning information of the photovoltaic equipment, and improving the accuracy and foresight of temperature anomaly prediction of the photovoltaic equipment.

For training a perception algorithm to detect an emergency vehicle, respective audio datasets are received from two microphones and respective spectrograms are generated. At least one interaural difference map is generated based on the spectrograms, audio source localization data is generated, which specifies a number of audio sources in respective grid cells of a spatial grid, by applying a CRNN to first input data containing the spectrograms and the least one interaural difference map. An image is received from a camera and output data comprising a bounding box for the emergency vehicle is predicted by applying at least one further ANN to second input data containing the image and the spectrograms. Network parameters are adapted depending on the output data and the audio source localization data.

The invention discloses a current sensor pin bending processing control system and processing equipment, and relates to the technical field of current sensor pin bending processing, and the current sensor pin bending processing control system comprises the following steps: fusing and collecting three types of disturbance parameters with different physical dimensions and response mechanisms, namely bending acceleration ZWa, pin surface contact impedance disturbance ZBd and friction sound spectrum response ZMc; the sensing coverage range of the system on the multi-source abnormal phenomenon in the machining process is remarkably expanded, the recognition blind area of a traditional system depending on a single force or displacement signal is made up, and the sensitivity to non-dominant disturbance characteristics is improved. Non-linear interaction modeling is carried out on the three disturbance factors through the perturbation coupling strength index psi, and the system not only can recognize single-variable fluctuation, but also can capture the co-evolution trend among variables. According to the structure, the system has the capability of recognizing disturbance structural imbalance and coupling activation trend in advance, and has higher fault perspectiveness and active response capability.

The invention relates to the technical field of livestock and poultry product safety detection, in particular to a temperature-controllable crushing device for livestock and poultry product detection, which is characterized in that sound wave signals generated in a homogenizing process are acquired through a sound sensor arranged outside a detection cup, and spectrum drift and energy attenuation characteristics of the sound wave signals are analyzed by adopting a sound spectrum temperature inversion algorithm; inverting the temperature and distribution of the sample in real time; and meanwhile, a cooling air curtain is formed through nitrogen injection, the gas flow rate and the injection angle are intelligently adjusted through a control unit according to a temperature inversion result, directional focusing cooling of a local overheating area is achieved, and a closed-loop temperature control system driven by acoustic signals is formed. According to the device and the method, cross contamination and sensor adsorption effect caused by contact type temperature measurement are thoroughly avoided, the problems that the sensor is easy to pollute and difficult to maintain in a high-humidity and fat-rich environment are solved, heat-sensitive veterinary drug decomposition caused by friction temperature rise in the homogenizing process is effectively inhibited, and the accuracy, reliability and repeatability of veterinary drug residue detection are remarkably improved.

The invention relates to the field of mechanical automation, in particular to a drilling machine intelligent operation system and a deviation correction control method.According to the system, through multi-dimensional signal fusion and dynamic risk assessment, drill jamming recognition and disposal response time is remarkably shortened, and efficiency loss caused by traditional shutdown protection is avoided; the imaging and obstacle positioning technology is adopted to replace manual drill lifting inspection, and the no-load operation time and mechanical loss of equipment are reduced. The multi-physical field signal fusion technology breaks through the sensing limitation of a single sensor, and is beneficial to accurately identifying underground obstacle space distribution and geological interface mutation features; acoustic spectrum and thermal field distribution analysis can enhance the ability to analyze drill bit-stratum interaction dynamics, and multi-dimensional input is provided for intelligent decision making. And dynamic optimization of a control strategy is realized in cooperation with a virtual-real interaction mechanism to adapt to demand changes of different stratum conditions and construction stages.

The invention relates to the technical field of data processing, in particular to an energy efficiency optimization method and system for a wind-liquid homologous data center, and the method comprises the steps: obtaining an acoustic spectrum signal during the operation of a cooling system through a sensor, and obtaining an original acoustic spectrum data set; performing frequency domain conversion on the original acoustic spectrum data set to obtain a frequency band distribution feature set; acoustic features are extracted according to the frequency band distribution feature set, and the system energy consumption state is determined in combination with a pre-established mapping model; judging the mode switching requirement of the cooling system; predicting an acoustic characteristic change trend according to a mode switching demand, and determining a mode switching time point; adjusting operation parameters of the cooling system according to the mode switching time point, and executing mode switching to obtain an updated system operation state; and updating the mapping model according to the updated system operation state to obtain an optimized energy consumption state prediction model. According to the invention, the problem of low precision of energy efficiency optimization of the data center is solved, and the precision of energy efficiency optimization is improved.

The invention relates to the technical field of unmanned aerial vehicle detection and tracking, in particular to a low-altitude unmanned aerial vehicle positioning and tracking method and system. The method comprises the following steps: acquiring sound spectrum time sequence data and radar plot data of an unmanned aerial vehicle; extracting a first sound spectrum characteristic parameter and a second sound spectrum characteristic parameter; extracting a first radar characteristic parameter and a second radar characteristic parameter; determining initial position estimation of the unmanned aerial vehicle through a first judgment condition; determining optimized position estimation of the unmanned aerial vehicle through a second judgment condition; and tracking the unmanned aerial vehicle by using an improved particle filtering algorithm to obtain a final position and a tracking trajectory of the unmanned aerial vehicle. According to the invention, through heterogeneous sensor data fusion, environment adaptive feature extraction, multistage intelligent judgment optimization and improved particle filter tracking, the precision, robustness and real-time performance of unmanned aerial vehicle detection in a complex environment are effectively improved, and the regional monitoring and countering capability is significantly enhanced.

The invention discloses a liquid metal bearing fault monitoring method and device, and belongs to the technical field of fault diagnosis. The method comprises the following steps: acquiring an acoustic signal during operation of a liquid metal bearing, processing and analyzing the acoustic signal, obtaining a plurality of singular spectrum components through a singular spectrum decomposition SSD algorithm and singular value decomposition, calculating a kurtosis value of each singular spectrum component, screening signal components according to the kurtosis values, reconstructing the signal, performing envelope demodulation on the reconstructed signal, and obtaining an envelope demodulation result of the liquid metal bearing. Feature extraction of the bearing fault is realized; and iteratively optimizing support vector machine (SVM) algorithm parameters by adopting a grey wolf optimization GWO algorithm, constructing a GWO-SVM algorithm, training the GWO-SVM algorithm through a training sample to generate a classification recognition model, and inputting a test sample to output a fault classification result. According to the liquid metal bearing fault monitoring method and device based on acoustic spectrum noise reduction analysis provided by the invention, the accuracy and efficiency of fault monitoring are improved, and the cost and risk of manual inspection are reduced.

The invention relates to the technical field of fire-fighting management and control, and particularly discloses a water level detection and fire-fighting identification method and system based on automatic control, and the method comprises the steps: obtaining the temperature and acoustic multi-modal data of a fire-fighting pipe network, building a pipeline temperature gradient model, and predicting the temperature distribution of each section through combining the environment and the medium temperature; and extracting ice crystal nucleation and fluid anomaly characteristics at low temperature. Based on the temperature gradient, the acoustic spectrum, the water level fluctuation rate and historical data, a multi-parameter fusion congelation risk assessment model is constructed, and risk indexes are calculated and graded; constructing a congelation risk and firelight early warning linkage control strategy, and adaptively adjusting a firelight detector threshold value and a sampling frequency according to a risk level; when the risk reaches an early warning value, partitioned heating is implemented, a temperature recovery curve and water level stability are collected in real time as feedback, model iterative optimization and closed-loop control are achieved, early warning accuracy is improved, energy consumption is reduced, and reliable operation of a fire fighting system is guaranteed.

The invention discloses a defibrillation decision-making method, device and equipment and a computer readable storage medium, and the defibrillation decision-making method comprises the steps: obtaining an electrocardiosignal, and carrying out the filtering and denoising of the electrocardiosignal, and obtaining an electrocardiosignal sequence; performing frequency domain analysis on the obtained electrocardiosignal sequence, and calculating frequency domain characteristic indexes including a sound spectrum flux, an optimized amplitude spectrum area and a sum of average amplitude difference functions; and according to the calculated frequency domain characteristic indexes, a comprehensive decision-making mechanism is adopted to carry out comprehensive decision-making to obtain a decision-making result of a defibrillation state, so that the problems that the defibrillation opportunity is difficult to grasp and the defibrillation effectiveness is limited are solved.

An embodiment of the present application provides a method, apparatus, device, and storage medium for audio classification, which relates to the field of artificial intelligence technology. The method for audio classification includes: obtaining the spectrogram features of an audio file; using a neural network model to obtain the first time-domain features of the spectrogram features; grouping the first time-domain features to obtain the second time-domain features of the spectrogram features, where the feature vector of each dimension of the second time-domain features represents the time-domain feature vector of each audio segment in the audio file; inputting the second time-domain features into a classification model to obtain the first feature labels of each audio segment in the audio file; and determining the type of the audio file according to the first feature labels of each audio segment. According to the local information of the audio segments in the audio file, the embodiment of the present application classifies the audio file, can pay attention to finer-grained local information, and helps to improve the accuracy of audio file classification.

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 provides an electric power circuit fault point positioning method and device based on voiceprint recognition, and relates to the technical field of electric power circuit detection, and the method comprises the steps: laying a voiceprint sensor for collecting a sound signal; preprocessing the first sound signal at the first position and then decomposing the first sound signal to obtain a first sound spectrum; traversing the first sound spectrum in a power line fault sound spectrum database to obtain a first most matched fault type corresponding to the first most matched sound spectrum; and when a first matching coefficient of the two positions accords with a preset matching threshold value, a first fault early warning instruction is sent, and fault early warning is carried out on the first position. The technical problem that the fault point positioning precision and the fault type identification accuracy are low due to the fact that the prior art depends on electrical signal perception and is easily influenced by electromagnetic interference is solved, and the technical effects that the fault positioning precision and the fault type identification capability are improved, and the fault early warning accuracy is remarkably improved are achieved.

Disclosed are integrated DFOS / DAS systems, methods, and structures that employ a large-scale pretrained recognition model we refer to as an “acoustic-language model”, which is pretrained with natural-language supervision (“contrastive language-audio pretraining”. The acoustic-language model comprises two primary components: an acoustic encoder and a text encoder. These encoders are pretrained using a cross-modal approach on a vast dataset of acoustic features (such as images created from log Mel spectrograms) and their corresponding textual captions. When acoustic features and / or languages are input into their respective encoders within the model, they generate corresponding embedding vectors. Both embedding vectors are then linked in a joint multimodal space using linear projections. The acoustic classification tasks using this model are executed by assessing the similarity between the acoustic and language embedding vectors, essentially evaluating the maximum similarity between the acoustic features and the events described in a specific language.

Embodiments of the present disclosure provide a singing evaluation method and device, a storage medium and a computing device. The method comprises: performing feature extraction on obtained singing audio to obtain a sound spectrum feature of the singing audio; inputting the sound spectrum feature into a multi-scale network model for calculation; wherein the multi-scale network model comprises a model based on a convolutional neural network, and a convolutional layer in the multi-scale network model comprises convolutional kernels corresponding to multiple time scales; and determining a final evaluation result based on a probability distribution of different evaluation results output by the multi-scale network model.

The present application relates to the technical field of artificial intelligence, and more particularly to an AI interaction system for ornithological scientific research, which comprises a collection unit, a trigger determination unit, a path determination unit, a dominant determination unit, a prompt generation unit and an adjustment unit. The present application constructs a multi-level behavior determination process through joint analysis of bird song spectrum, image motion characteristics, flight trajectory and environmental factors, and through statistical analysis of the time distribution and spatial distribution of these behavior indicators within a preset period, and correlation analysis with environmental factors such as temperature, humidity, wind speed and illumination, the influence of environmental changes on bird activity frequency and spatial aggregation degree can be identified, so that the acoustic trigger threshold can be dynamically adjusted, so that the system can continuously maintain the sensitive recognition ability of bird activity with the change of the habitat environment, and then realize high-reliable behavior recognition and interaction prompt for ornithological scientific research.

The invention discloses a method, a system, equipment, a medium and a program for intelligently diagnosing faults of multiple parts of a wind power generator cabin, and belongs to the technical field of wind power generation. The method comprises the following steps: collecting a multi-source acoustic signal in a cabin of the wind power generator, preprocessing the multi-source acoustic signal to obtain a preprocessed voiceprint signal, and generating a spectrogram; performing cross-domain feature extraction on the pre-processed voiceprint signal, performing feature dimension reduction and redundancy elimination through principal component analysis, and constructing a voiceprint feature vector; inputting the spectrogram and the voiceprint feature vector into a CNN-LSTM fusion diagnosis model, and performing fault diagnosis to obtain a fault diagnosis result; and matching the fault diagnosis result with the fault-voiceprint knowledge graph to generate a decision suggestion. According to the method, through non-contact voiceprint monitoring and multi-dimensional analysis, high-precision and real-time diagnosis and active decision-making of concurrent faults of multiple components of the wind power generator cabin are realized, and the defects of complex deployment and sound signal interference resistance of traditional vibration monitoring are effectively overcome.