55 results about "Stochastic resonance" patented technology

The invention discloses a low-illumination image quality improvement method based on self-adaptive stochastic resonance. The method is specifically implemented according to the following steps: step 1, synchronously acquiring a grayscale image and a color original image corresponding to the same image through two channels; step 2, preprocessing the grayscale image and the color original image acquired in the step 1; 3, self-adaptive stochastic resonance parameter calculation is carried out; step 4, performing dual-channel adaptive stochastic resonance processing on the preprocessed grayscale image and the preprocessed color original image according to adaptive stochastic resonance parameters to obtain an enhanced image; and 5, carrying out brightness component fusion on the enhanced image, and then carrying out image reconstruction and post-processing. According to the invention, the problem of poor image quality after low-illumination image enhancement in the prior art is solved.

The invention discloses a phase modifier bearing fault diagnosis method, which belongs to the technical field of power equipment fault diagnosis, and comprises the following steps of: analyzing independent components, performing blind source separation on a received phase modifier bearing vibration signal, and extracting three types of independent source signals of impact, abrasion and noise; the gradient driving window length is self-adaptive, and the window length of short-time Fourier transform is dynamically adjusted based on the instantaneous frequency gradient of the independent source signal; wavelet packet frequency band energy screening: performing wavelet packet decomposition on the signal after window length adaptive processing, screening a fault characteristic frequency band based on an energy contribution rate, and reconstructing the signal; and enhancing stochastic resonance, and inputting the reconstructed signal into a stochastic resonance system. According to the method, independent component analysis, gradient driving window length self-adaption, wavelet packet frequency band energy screening and stochastic resonance enhanced fourth-order diagnosis chain are constructed, so that multi-stage cooperative processing of phase modifier bearing faults is realized, and the technical problems of time-frequency resolution contradiction, insufficient feature decoupling and weak generalization ability are effectively solved.

The invention discloses an underwater target signal enhancement method based on bistable stochastic resonance and a particle swarm optimization algorithm, and the method comprises the steps: S1, collecting a simulation perception electric signal corresponding to a flow field caused by the disturbance of a propeller of an underwater vehicle through a bionic side line perception sensor based on friction nanometer power generation; s2, preprocessing the simulated sensing electric signal to obtain a preprocessed signal sequence; and S3, constructing a target signal enhancement strategy based on bistable stochastic resonance and a particle swarm optimization algorithm, and performing signal enhancement processing on the preprocessed signal sequence based on the target signal enhancement strategy to obtain a target enhanced signal. According to the invention, the problem that the existing method cannot effectively realize underwater target signal enhancement is solved.

This invention discloses a method for enhancing weak underwater signals using stochastic resonance, belonging to the field of underwater acoustic engineering and signal processing technology. It is used for enhancing weak underwater signals, including constructing and preprocessing an integrated detection and communication signal; constructing a linear potential well wall piecewise tristable stochastic resonance system; inputting the preprocessing results into the linear potential well wall piecewise tristable stochastic resonance system; optimizing the parameters of the linear potential well wall piecewise tristable stochastic resonance system using an improved moss growth optimization algorithm; numerically solving the system's motion equations using the fourth-order Runge-Kutta method to obtain the enhanced output signal; and performing frequency recovery on the enhanced output signal to generate an output signal with a complete linear frequency modulation structure. This invention fundamentally solves the problems of output saturation and difficulty in improving the signal-to-noise ratio in traditional stochastic resonance systems under strong noise environments through the linear potential well wall piecewise tristable structure, significantly improving the system's adaptability to complex underwater noise.

The invention discloses a method and a system for detecting weak signals and harmonic waves based on stochastic resonance of compression enhancement. The system comprises a vacuum system, an imprisoned field, a laser system, a fluorescence acquisition system and an external signal, the vacuum system comprises a vacuum cavity, an ion pump and a getter pump, and a calcium atom furnace and a surface electrode ion trap for trapping calcium ions are placed in the vacuum cavity; the trapping field comprises a radio frequency trapping field and a direct current trapping field, and the radio frequency trapping field is composed of a radio frequency signal source, a power amplifier and a spiral resonant cavity; the direct-current trapping field comprises a direct-current high-voltage power supply, a 6733 board card and an amplification chip. The fluorescence acquisition system comprises an imaging mirror, a CCD (Charge Coupled Device) camera and a photomultiplier; the external signal comprises a driving signal, a compression signal and a to-be-tested signal. According to the method, the compressed signal is used for replacing external random noise, ions are triggered to generate periodic phase change along with the to-be-detected signal, the to-be-detected signal is optimally matched with the bistable system by changing the strength of the compressed signal, and the signal-to-noise ratio and harmonic detection are improved to the maximum extent.

The invention provides a method for extracting features of ultrasonic waves generated by a typical discharge phenomenon in a complex working condition environment, and belongs to the technical field of power equipment.The method comprises the steps that a transient feature vector is extracted by conducting synchronous compression wavelet transform on a signal component, the transient feature vector is input into a waveform propagation enhancement model, and an enhanced feature vector is output; stochastic resonance amplification is carried out on the enhanced feature vector to establish a weak feature vector, the weak feature vector and the transient feature vector are fused to establish an equivalent discharge intensity matrix, a discharge type discrimination index is calculated, the discharge type is determined according to the discrimination index, and a propagation speed parameter of a multi-path delay correction matrix is adaptively adjusted. And on the basis of the updated propagation path analysis result, a time difference of arrival matrix and a maximum likelihood estimation method are utilized to calculate the spatial position coordinates of the discharge source, and the problem of low discharge ultrasonic feature extraction accuracy caused by multi-medium propagation path interference in a complex working condition environment is solved.

The application discloses a fault positioning method and device, electronic equipment, storage medium and program product; the method comprises the following steps: when it is detected that a traveling wave analysis starting condition is met, a fault traveling wave signal set to be analyzed of two measuring points of a line is acquired, the fault traveling wave signal set comprises fault traveling wave signals collected in different sampling periods, and the traveling wave analysis starting condition is determined according to a gain function of a power frequency signal; for each fault traveling wave signal set, a random resonance system potential function is constructed according to the fault traveling wave signal set, system resonance is performed on the fault traveling wave signal according to the random resonance system potential function, and a traveling wave signal to be extracted is obtained; a wave head is extracted from the traveling wave signal to be extracted, and a traveling wave arrival time corresponding to the measuring point is recorded; and fault positioning is performed according to the traveling wave arrival time corresponding to each measuring point and the total length of the line, and a fault position is determined, so that the problems of inaccurate fault traveling wave signal collection and inaccurate wave head extraction are solved.

The application provides a weak non-stationary signal processing method and system, and relates to the technical field of signal processing, and comprises the following steps: obtaining a weak non-stationary signal; estimating the signal frequency range of the weak non-stationary signal, and estimating the noise intensity according to the estimated signal frequency range; constructing a stochastic resonance model based on a Gaussian mixed potential function, and adaptively optimizing the Gaussian mixed potential function according to the estimated noise intensity, so as to obtain an optimal stochastic resonance model, with the maximum spectral amplification factor of the stochastic resonance model as the target; constructing a parameterized demodulation operator to demodulate the weak non-stationary signal, inputting the demodulated signal into the optimal stochastic resonance model for enhancement, and finally obtaining the enhanced non-stationary signal through inverse demodulation; and the application is based on Gaussian mixed potential function stochastic resonance, and solves the bottleneck problem of weak non-stationary signal processing through a closed-loop process of noise intensity estimation, potential function optimization, frequency estimation and signal enhancement.

The application relates to an underwater optical communication receiving signal detection method. The method comprises the following steps: pre-processing a receiving signal disturbed by turbulence by using secondary sampling to obtain a small parameter signal; using a Kalman filtering algorithm to obtain the envelope of the receiving signal disturbed by turbulence by processing the small parameter signal; processing the envelope of the receiving signal disturbed by turbulence according to a signal equalization formula; establishing an adaptive stochastic resonance system model, and taking the processed signal as the input signal of the adaptive stochastic resonance system model; using a multi-strategy fusion particle algorithm to optimize the system parameters in the stochastic resonance system model by taking the system output signal-to-noise ratio as a target function; judging whether the iteration of the multi-strategy fusion particle algorithm is terminated, and if the iteration is terminated, outputting the optimal system parameters corresponding to the maximum value of the target function; and inputting the optimal system parameters into the adaptive stochastic resonance system model. The application can realize adaptive stochastic resonance detection.

The invention relates to a training method of a frequency difference estimation model, belongs to the technical field of radiation source signal positioning, and solves the problem that the prediction precision of an existing frequency difference estimation model on a signal frequency difference is poor. The training method comprises the steps that multiple tests are carried out by changing the positions of radiation source signals, two observation stations are used for receiving signals in each test, the signals, received by the two observation stations, of the radiation source at the same position serve as a set of signals, and the frequency difference of each set of signals is marked; performing signal enhancement processing on each group of signals by using a preset optimal stochastic resonance system, and taking each group of signals after signal enhancement processing and the frequency difference of each group of signals as a training sample to obtain a training set of a frequency difference estimation model; training the frequency difference estimation model based on the training set to obtain a trained frequency difference estimation model; wherein the frequency difference estimation model adopts a convolutional neural network. And the frequency difference prediction precision of the frequency difference estimation model is improved.

The application discloses a method and system for detecting weak signals and harmonics based on compression-enhanced stochastic resonance, comprising a vacuum system, a trapped field, a laser system, a fluorescence collection system and an external signal; the vacuum system comprises a vacuum cavity, an ion pump and a suction pump, a calcium atom furnace and a surface electrode ion trap for trapping calcium ions are arranged in the vacuum cavity; the trapped field comprises a radio frequency trapped field and a direct current trapped field, the radio frequency trapped field is composed of a radio frequency signal source, a power amplifier and a spiral resonant cavity; the direct current trapped field comprises a direct current high-voltage power supply, a 6733 board card and an amplification chip; the fluorescence collection system comprises an imaging mirror, a CCD camera and a photomultiplier tube; the external signal comprises a driving signal, a compression signal and a signal to be detected. The application uses the compression signal to replace the external random noise, triggers the ion to have periodic phase change with the signal to be detected, changes the intensity of the compression signal, makes the signal to be detected and the bistable system reach the best matching, and maximizes the signal-to-noise ratio and the detection of harmonics.

The invention relates to a variable-speed bearing fault feature extraction method and device, relates to the technical field of bearing fault diagnosis, and aims to solve the problem that the variable-speed bearing fault feature extraction precision is not high. The method comprises the steps that a particle swarm optimization algorithm is introduced into a multi-scale adaptive morphological filtering operator, key parameters of the adaptive morphological filtering operator are obtained, filtering signals of all scales are obtained through calculation according to the multi-scale adaptive morphological filtering operator and the key parameters, and the filtering signals of all the scales are added to obtain a primary filtering signal; introducing a particle swarm optimization algorithm into an adaptive stochastic resonance algorithm to obtain a system cooperative adjustment parameter of the adaptive stochastic resonance algorithm when the signal-to-noise ratio is maximum, and inputting the primary filtering signal into the adaptive stochastic resonance algorithm to obtain a secondary filtering signal; and calculating the secondary filtering signal by using an angular domain resampling algorithm to obtain an order spectrogram containing bearing fault features. The method has the effect of improving the fault feature extraction precision of the variable-speed bearing.

The application relates to a bistable SR system and a low signal-to-noise ratio multi-frequency sound signal detection device and method. By introducing an optical fiber device, the closed-loop regulation of the optical fiber grating is utilized to realize the random resonance phenomenon based on the bistability, and the low signal-to-noise ratio signal can be effectively extracted under the strong noise background. Secondly, by introducing a carrier signal, the difference frequency signal is obtained by mixing with the external signal, and the multi-frequency signal sensing is realized under the premise of fixing the detection range of the difference frequency signal. In addition, by introducing the array structure of the optical fiber hydrophone, the spatial gain is formed, and the high-sensitivity detection effect can be realized. Finally, the frequency distribution and the relative signal-to-noise ratio intensity size distribution of the obtained target sound signal can be used as the input of the neural network model, the intelligent identification of the low signal-to-noise ratio target sound target is realized, and the detection and early warning of the underwater target can be maximally realized.

The application belongs to the technical field of photovoltaic electrical fault detection, and particularly discloses a research method for detecting features of a direct-current fault arc of a photovoltaic system based on adaptive morphological enhancement. In order to ensure the reliability of the system, the fault arc is accurately detected and the detection features are enhanced. The enhancement effect of morphological on the detection features of the direct-current fault arc is studied, and an adaptive morphological algorithm is proposed. The particle swarm algorithm is combined with signal features to dynamically optimize the structure element, improve the signal signal-to-noise ratio, enhance the detection features, and compared with the stochastic resonance and wavelet analysis algorithm, so that the good feature enhancement effect of the application is verified, and the normal and fault arcs can be effectively distinguished. Finally, the features are decided by combining the support vector machine, and a detection accuracy of 96.23% is obtained. The adaptive morphological algorithm can effectively enhance the detection features of the fault arc, is suitable for adaptive optimization in different experimental scenes, and is helpful for accurately and effectively detecting the direct-current fault arc.

This invention relates to the field of elevator technology, and more particularly to an anomaly detection method for elevator polyurethane buffers based on millimeter-wave radar. The method includes: acquiring echo signals within a preset time window after the polyurethane buffer's compression and rebound by millimeter-wave radar; determining the target distance unit and generating a vibration displacement signal in the radar's line-of-sight direction; identifying intervals where the overall vibration envelope shows a decaying trend and extracting the damping response signal; implementing bistable stochastic resonance enhancement and variational mode decomposition to identify the target modal signal; calculating damping characteristic indices using the target envelope amplitude and stiffness characteristic indices using the target phase to construct a health feature vector; and determining whether the obtained health feature vector lies within the constructed health reference space. Using non-contact millimeter-wave radar to acquire echo signals during the polyurethane buffer's rebound process allows for precise capture of phase changes caused by minute movements or vibrations of the target, resulting in high detection efficiency and safety.

The application provides a kind of self-adapting random resonance weak fault detection method based on kurtosis optimization, belongs to the technical field of fault detection, including: collecting transient signal on cable line and preprocessing;Constructing bistable random resonance model;With the kurtosis of output signal as fitness function, the potential well parameters of bistable random resonance model are optimized based on particle swarm optimization algorithm;The optimized potential well parameters are substituted into the bistable random resonance model, and the preprocessed transient signal is processed to obtain the enhanced output signal for fault detection.The beneficial effect is: based on the classical bistable model to construct the random resonance system, the kurtosis of output signal is used as the fitness function of particle swarm optimization algorithm, the potential well parameters are adaptively optimized, the potential well parameters are automatically converged to the optimal state, the high impedance grounding fault traveling wave signal can be extracted and enhanced from strong noise background, and the detection sensitivity and reliability are improved.

The invention relates to the technical field of hydraulic systems, in particular to a cleanliness detection method and system of a hydraulic system. The method comprises the following steps: acquiring monitoring signals of the hydraulic system, wherein the monitoring signals comprise an initial photoelectric signal of a photoelectric sensor for detecting oil and a mechanical vibration signal at the same position; the monitoring signal is divided into a plurality of analysis windows, the actual pulse width factor of the mechanical vibration signal in each analysis window is calculated, the potential well width parameter is adjusted through the actual pulse width factor to obtain a self-adaptive potential well width parameter, and the self-adaptive potential well width parameter is inversely proportional to the square of the actual pulse width factor; and processing the initial photoelectric signal by using a stochastic resonance system based on the adaptive potential well width parameter, and performing cleanliness detection based on the processed signal. The method has the effect of improving the oil cleanliness detection precision of the hydraulic system.

The present application provides a device, system and method for detecting weak electromagnetic signals of a sample, comprising a driving module, N Helmholtz coils, a low-temperature container, two SQUID modules, a first signal processing module and a second signal processing module; N is a natural number greater than 1; the driving module provides driving signals for each Helmholtz coil, each SQUID module includes a SQUID galvanometer and a gradiometer; the input end of each SQUID galvanometer is connected to each gradiometer; the input end of the first signal processing module is connected to the output end of each SQUID galvanometer, and outputs a first processing signal to the second signal processing module; the second signal processing module determines whether the processing signal produces random resonance. The present application can realize synchronous detection, that is, the sample of the substance to be tested dissolved in the solvent and the sample containing pure solvent are tested synchronously at the same time, further avoiding the error caused by the time difference caused by separate testing.

This invention discloses an airborne magnetic anomaly detection method based on a stochastic resonance system, comprising the following steps: S1: Acquiring geomagnetic field data. The data is processed using airborne magnetic compensation technology. S2: Establishing a stochastic resonance detection system and calculating a binary hypothesis detection statistic. The binary hypothesis detection statistic is the standard deviation of the output of the stochastic resonance system. S3: Calculating a threshold and detecting the target. The threshold is calculated using the unit averaging method. The presence of a target is determined by comparing the binary hypothesis detection statistic with the threshold. A target is considered to exist when the binary hypothesis detection statistic of any subsystem in the stochastic resonance system is greater than the corresponding threshold; the measurement is considered pure noise only when both subsystem statistics are less than the threshold. This method addresses the problems of low signal-to-noise ratio and the severe influence of noise type on correlation matching detection methods during magnetic anomaly detection in complex environments.

The present application belongs to one-dimensional signal processing technical field, and particularly relates to a Duffing oscillator weak signal enhancement method with automatic parameter optimization. The present application takes a double-stage parallel Duffing random resonance system as a mixed signal processing model, uses a quantum particle swarm algorithm with outstanding performance in optimization solution as a parameter optimization algorithm, converts the double-stage parallel random resonance parameter adjustment problem into a multi-parameter global optimization problem, and performs signal enhancement processing on the signal to be enhanced input into the system according to the best parameters, and outputs the enhanced signal. The present application uses parameter self-adaptive optimization to quickly obtain the best parameters of the double-stage parallel Duffing random resonance system, greatly improves the efficiency and precision of the system best parameter setting, and can quickly and high-quality output the enhanced useful signal.

This invention provides a method for constructing a small bowel CT image recognition and segmentation network model and a recognition method. The construction method includes the following steps: acquiring several small bowel CT images showing lesions and constructing an image dataset, dividing the dataset into a training set and a test set proportionally; enhancing the image dataset in the above steps using dynamic stochastic resonance enhancement; constructing a U-Net network model and training the network model using the enhanced training set obtained in the above steps; testing the network model using the enhanced test set obtained in the above steps, and completing the model construction. This invention proposes a method for enhancing CT images based on dynamic stochastic resonance, which can effectively enhance useful signals in the original image. Since medical image processing is mainly used to assist doctors in diagnosis, the input is an image, and the output is also an image that is easier for doctors to use. This invention establishes a U-shaped neural network that meets this requirement.

The invention discloses a signal-to-noise ratio enhancement method, a balance detection method and a circuit, and the method comprises the steps: adjusting the gain and reference threshold voltage according to an output signal of an analog-to-digital converter, enabling the mean value of the threshold voltage to be closest to the mean value of a signal after gain processing, enabling the peak-to-peak value of the signal after gain processing to be larger than a threshold voltage difference value, and achieving the balance detection of the signal-to-noise ratio. Under the condition, noise carried in the signal after gain processing is used for driving comparison output to jump between a high level and a low level, a stochastic resonance effect is formed, and therefore the signal-to-noise ratio of the stochastic resonance signal is greatly increased.

This invention discloses a bearing fault diagnosis method based on GA-VMD and adaptive stochastic resonance, belonging to the field of early fault diagnosis of rotating machinery. The implementation method is as follows: First, the original rolling bearing vibration signal is acquired. Second, using envelope entropy as the comprehensive objective function, a genetic algorithm is used to search for the minimum value of the comprehensive objective function, determining the optimal combination of the penalty parameter α and the number of modes k in the variational mode decomposition algorithm. Third, the original bearing signal is initially denoised using the optimized variational mode decomposition algorithm. Fourth, using the signal-to-noise ratio of the bearing signal as the comprehensive objective function, a quantum particle swarm optimization algorithm is used to search for the maximum value of the comprehensive objective function, determining the optimal combination of the nonlinear system potential function parameters a and b and the damping coefficient μ, obtaining an optimal parameter-adaptive stochastic resonance system. Fifth, this stochastic resonance system is used to perform stochastic resonance on the bearing vibration signal to improve the signal-to-noise ratio of the bearing signal. Sixth, spectral analysis is performed on the stochastic resonance output signal to achieve accurate extraction of weak fault features and accurate fault identification.

The invention discloses a water pump cavitation diagnosis method based on nonlinear chaotic characteristics of vibration signals, and belongs to the technical field of water pump cavitation diagnosis. Performing phase-space reconstruction on the high-frequency vibration signal; extracting a nonlinear chaotic feature set from the reconstructed phase space, and judging whether the cavitation is early-stage cavitation or not; according to the separation degree index of each feature in the normal state and the cavitation state, calculating an adaptive weight, and carrying out weighted fusion to obtain a comprehensive feature; and inputting the comprehensive features into a bistable stochastic resonance classifier, enhancing the features through a kinetic equation, and outputting cavitation state classification. According to the method, high-frequency vibration signals larger than or equal to 200 kHz are collected, phase-space reconstruction and nonlinear chaos feature extraction are combined, a bistable stochastic resonance classifier is matched, the early-stage cavitation and the cavitation germination state of the water pump are accurately recognized, millisecond-level real-time diagnosis is achieved through edge calculation, and the accuracy of the early-stage cavitation and the cavitation germination state of the water pump is improved. The problems that a traditional method is insensitive to nonlinear characteristics, depends on cloud computing power and is large in delay are solved.

The invention relates to the technical field of machine vision and intelligent safety monitoring, in particular to a high oxygen partial pressure environment personnel state monitoring system based on machine vision. A data acquisition module, a feature processing module and a reliability analysis module; the system determines a stimulated stochastic resonance enhancement coefficient by acquiring environment pressure, oxygen partial pressure and image features; the core of the method is that according to the coefficient and the time sequence parameter, the interruption decoupling probability of the environmental noise and the personnel semantic primitive is calculated, so that the effective recognition intensity is determined, and the personnel consciousness state is evaluated; a stimulated stochastic resonance mechanism is utilized, resonance is induced by actively injecting auxiliary noise, the detection rate of weak signals is improved, weak sign loss caused by traditional filtering is avoided, and the detection limit in an extreme signal-to-noise ratio environment is remarkably improved.

The invention discloses a weak signal detection method based on an improved under-damping unsaturated bistable stochastic resonance model, and the method comprises the steps: firstly collecting an original weak signal of a detected rolling bearing, carrying out the secondary sampling of the original weak signal, enabling a target frequency to be mapped to an adiabatic approximate working interval according to a proportionality coefficient, and enabling the target frequency to be converted into an adiabatic approximate working interval; obtaining a pre-processed sampling signal; thirdly, constructing an improved under-damping unsaturated bistable stochastic resonance model, substituting the sampling signal into the improved under-damping unsaturated bistable stochastic resonance model, and then performing optimization solution through a particle swarm algorithm to obtain optimal model parameters; and finally, substituting the optimal model parameters into the improved under-damping unsaturated bistable stochastic resonance model to obtain a model output state signal, and performing fast Fourier transform on the state signal to extract the target characteristic frequency to obtain the model output signal, namely the extracted weak signal.

The present application relates to the technical field of bearing fault signal detection, and particularly relates to a bearing fault diagnosis method based on adaptive fractional order stochastic resonance, which introduces a time delay feedback term into a fractional order stochastic resonance system in a stochastic resonance detection method, finds the structural optimal parameters of the stochastic resonance system by improving a cuckoo algorithm, substitutes the structural optimal parameters into a time delay fractional order biased nonlinear overdamped stochastic resonance system, and finally extracts rolling bearing fault features and detects fault feature frequencies through time-frequency domain analysis on the signal processed by the stochastic resonance system, so that bearing fault diagnosis is realized through comparison. Compared with existing diagnosis methods based on stochastic resonance systems, the present application has a simple model, fewer algorithm parameters, fast convergence speed, and can accurately and efficiently detect bearing fault signals.

This invention relates to the field of stochastic resonance, specifically to a stochastic resonance simulation system, comprising: a stochastic resonance parameter generation module, used to match a corresponding stochastic resonance simulation parameter table to the characteristic parameters of the current frequency signal; a stochastic resonance simulation module, used to build a bistable system stochastic resonance model based on Simulink, and adaptively select the corresponding stochastic resonance simulation parameters from the stochastic resonance simulation parameter table to complete the stochastic resonance simulation; a virtual parameter module, which is a logic unit inserted into the bistable system stochastic resonance model that can directly obtain the corresponding results or information; and a stochastic resonance simulation circuit, used to perform denoising processing on the current frequency signal based on the optimal simulation parameters obtained by the stochastic resonance simulation module. This invention can adaptively and quickly find the most suitable stochastic resonance simulation parameters according to different frequency signal enhancement targets, while improving the accuracy of stochastic resonance simulation results.

The application discloses a rolling bearing fault recognition method based on stochastic resonance and a convolutional neural network, comprising the following steps: setting artificial fault points on inner and outer rings and rolling bodies of a rolling bearing, collecting bearing vibration signals of the rolling bearing under different rotating speeds, constructing an adaptive variable scale stochastic resonance detection system based on an improved signal-to-noise ratio, performing uniform scale processing on a horizontal axis on spectrum graphs respectively, composing an original image set by using the processed spectrum graphs, marking and classifying spectrum graphs in the original image set based on the artificial fault points, dividing the original image set after marking and classifying into a training set and a test set, building a convolutional neural network model and performing training and testing, determining parameters of the convolutional neural network model according to a test result and representing the parameters as a recognition model, and performing fault recognition on acquired bearing vibration signals to be recognized based on the recognition model and acquiring a fault type. The application solves the problem of weak fault recognition of the rolling bearing under strong noise and variable rotating speed working conditions.