2230 results about "De noise" patented technology

The invention discloses a data center intelligent operation and maintenance method and system based on an industrial internet of things, belongs to the field of image processing calculation, and aims to solve the problems of privacy protection deficiency, serious manual dependence and weak energy efficiency and safety. Full-link dynamic optimization is realized through an edge calculation layer and an adaptive clustering network, edge nodes are integrated with a wavelet transform noise filtering technology, dynamic denoising is performed on multi-band sensor signals of temperature, vibration, current and the like, heterogeneous device time sequence data are aligned in combination with a dynamic time warping algorithm, sampling frequency differences of multiple devices are eliminated, and the multi-band sensor signals are obtained. A weight formula is put forward, a cluster head is dynamically elected based on node residual energy, data similarity and a load coefficient, a transmission path of a high-delay area is optimized in combination with a multi-hop routing protocol, the packet loss rate and the signal strength are evaluated in real time, redundant coding is started for a high-packet-loss link or the high-packet-loss link is switched to a low-power-consumption mode. And synchronizing the dynamic topology state with the digital twin platform through an open platform communication unified architecture protocol.

The invention provides a power grid fault waveform identification and intelligent relay protection rapid control method and system, and relates to the technical field of electric power, and the method comprises the steps: collecting the voltage and current waveform data of a power grid, carrying out the digital filtering and noise elimination processing, carrying out the Fourier transform and wavelet transform, extracting the time domain and frequency domain features, and constructing a combined feature vector. And a deep learning model including a convolutional neural network and a long-short-term memory network is used to identify and evaluate the type, position and amplitude of the power grid fault based on the feature vector database. And then, a multi-target particle swarm optimization algorithm for constructing a target weight matrix based on historical fault data is adopted to optimize fault isolation time, a power supply recovery rate and system stability to obtain a control strategy, and a relay protection device setting value is adjusted according to the control strategy to execute fault isolation and power grid topology reconstruction.

The invention discloses a composite power quality disturbance identification method based on a time-frequency feature fusion classification network. The method comprises the following steps: carrying out noise reduction processing on a noise-containing disturbance signal based on self-adaptive noise complete mode decomposition in combination with a wavelet threshold method; obtaining frequency domain data of the noise reduction signal based on fast Fourier transform; based on a TCN-Transformer deep network, deep feature extraction is carried out on the denoised signals in parallel from a time domain and a frequency domain; and carrying out weighted fusion on the extracted signal time domain and frequency domain features based on a cross attention mechanism to finish classification and identification. The method is simple in principle, the average recognition accuracy rate in a 30dB noise environment can reach 97.86%, and compared with other mainstream deep learning recognition models, the method shows better recognition precision, recognition stability and generalization ability.

The invention relates to a cloud edge cooperative computing framework and processing method for multi-modal data stream fusion processing, and the method comprises the following steps: S1, carrying out the noise suppression based on an original data stream collected by an edge computing node through employing an improved Wiener filtering algorithm, achieving the signal denoising through the adaptive threshold wavelet transformation, and obtaining a cloud edge data stream; and a timestamp alignment technology is utilized to solve the problem of time delay difference of multi-modal data, and a space-time alignment purified data stream is generated. Through combination of the improved Wiener filtering algorithm and the adaptive threshold wavelet transform, the noise suppression efficiency of the original data stream is significantly improved, the timestamp alignment technology effectively solves the time delay difference of the multi-modal data, the generation of the space-time alignment purified data stream ensures that the subsequent processing has a unified time sequence benchmark, and the efficiency of noise suppression of the original data stream is improved. The space-time attention fusion network adopts a collaborative architecture effect of a bidirectional gating circulation unit and a lightweight 3D convolutional network.

The invention discloses a water quality heavy metal pollution detection method and system based on Raman spectrum.The water quality heavy metal pollution detection method comprises the steps that water body Raman scattering light is collected in situ through a miniature optical fiber probe, and a continuous time sequence spectrum signal flow is generated; wavelet transform is combined with self-adaptive threshold setting, and high-frequency noise and effective spectral signals are separated; dynamically strengthening the characteristic peak of the target heavy metal through a frequency domain characteristic screening module, and inhibiting a water molecule interference peak at the same time; generating a cross-domain fusion feature vector; processing the fusion feature vector through a pre-trained heavy metal concentration prediction model, and outputting concentration prediction values of various heavy metals; a confidence score is dynamically calculated based on a deviation between a current predicted value and historical data distribution, and model parameter update and system calibration are automatically triggered. The method has the advantages that through acousto-optic signal cross-domain fusion and closed-loop self-calibration, the target peak is dynamically strengthened while water molecule interference is inhibited, and the real-time performance, the anti-interference performance and the prediction precision of heavy metal detection are improved.

The invention relates to the technical field of hydrogeological monitoring, in particular to a hydrogeological dynamic monitoring and analysis system and method based on big data. The method specifically comprises the steps that hydrogeological data are collected in real time, edge calculation is combined for data denoising, space-time calibration and error correction, variational mode decomposition, geologic model constraint and wavelet transform are adopted for eliminating noise and drift errors, storage and transmission efficiency are optimized through incremental coding and Huffman coding, and quadratic form calibration codes are generated to ensure data reliability; a hydrological evolution trend is predicted by fusing a space-time deep learning model and a self-attention mechanism, and adaptive anomaly detection is realized by combining DBSCAN clustering and prediction error analysis; and optimizing a water resource allocation strategy through reinforcement learning, hierarchically evaluating a risk level and generating a management scheme, and finally visually displaying monitoring data, an abnormal score and a decision result. According to the method, multi-source data fusion, space-time modeling and intelligent decision making are integrated, and hydrological monitoring precision and response efficiency are improved.

The invention relates to a remote sensing image change detection system and method based on multi-modal deep learning. According to the system, the detection precision is improved by constructing a twin network fusing a CNN, a Transform and an attention mechanism; a lightweight backbone network is constructed by adopting multi-size depth separable convolution, and noise is adaptively suppressed and feature expression is enhanced in combination with a soft thresholding channel space attention mechanism; harr wavelet transform downsampling is innovatively introduced to reserve high-frequency details, and traditional pooling operation is replaced to reduce information loss; the twinborn branch semantic deviation is relieved through cross-channel feature exchange, and a Transform codec is used for modeling a long-range dependency relationship; in the decoding stage, edge feature recovery is enhanced by adopting feature splicing and a progressive up-sampling strategy; according to the system, model parameters are reduced, meanwhile, the precision and the anti-interference capability of remote sensing building change detection are remarkably improved, and the system has the advantages of high efficiency and detail keeping.

The invention relates to the technical field of power system monitoring, and discloses an infrared thermal image feature fusion-based power transmission line abnormal area identification method, which comprises the following steps of: acquiring infrared thermal image data of a power transmission line, and preprocessing the data; the invention further provides an infrared thermal image feature fusion power transmission line abnormal area identification system, which comprises a data acquisition module, an infrared thermal image feature fusion module, an infrared thermal image feature fusion module, an infrared thermal image feature fusion module, an infrared thermal image feature fusion module and an infrared thermal image feature fusion module, and is characterized in that the data acquisition module is used for acquiring infrared thermal image data of a power transmission line; and the preprocessing module is used for preprocessing the collected infrared thermal image data. According to the method, through combination of multi-scale wavelet transform, nonlinear dynamic modeling, graph neural network optimization and reinforcement learning, the abnormal region identification precision is improved, the robustness and adaptive ability of the system in a complex environment are enhanced, and it is ensured that the abnormal region can still be efficiently and accurately identified under dynamic change and noise interference.

The invention provides a cluster network flow prediction method based on multi-scale time feature fusion, and belongs to the technical field of computer network flow prediction. The method comprises the following steps: determining a multi-index prediction sequence based on traffic load characteristics of cluster IP instances, and constructing a high-quality time sequence data set; fourier transform and discrete wavelet transform are used for time-frequency feature analysis, and noise filtering and data dimension reduction are completed; projecting sequences of different time granularities to a unified model dimension, performing one-dimensional channel convolution merging, inputting the merged sequences into a time encoder and a cross-channel encoder, and capturing cross-scale long-term time dependence and a coupling relationship between variables; in the loss function design, time domain and frequency domain loss are fused, double-domain error calculation is carried out on a prediction result and a label through Fourier transform, and the robustness of a model to non-stationary fluctuation is enhanced; and through linear layer decoding and reverse normalization processing, the abstract feature is converted into an actual flow prediction value. According to the invention, the precision and reliability of cluster network flow prediction are significantly improved.

The invention relates to a water quality real-time detection system which comprises the following modules: a multi-source sensing module which is based on a multi-parameter sensing array, adopts a self-adaptive sampling strategy, realizes sensor time sequence synchronization through a state estimation algorithm, completes water body multi-dimensional parameter acquisition in combination with a micro-fluidic chip, generates a multi-modal sensing data set, and transmits the multi-modal sensing data set to a data processing module; the multi-source sensing module comprises a multi-source sensing sub-module, a signal conditioning sub-module, a time sequence synchronization sub-module and an anomaly capture sub-module. The method has the advantages that through the synergistic effect of the adaptive sampling strategy and the state estimation algorithm, the multi-sensor time sequence synchronization precision is remarkably improved, the phase deviation problem caused by traditional fixed frequency sampling is effectively eliminated, the sliding window polynomial fitting is combined with the wavelet threshold de-noising technology, and the multi-sensor time sequence synchronization precision is improved. High-frequency noise interference is greatly suppressed on the premise that effective components of the signals are reserved, and meanwhile, the abnormal value detection accuracy is improved through a dynamic threshold mechanism.

The invention discloses a drilling multi-source sensing signal denoising method based on dynamic noise decoupling and a self-adaptive mode, and belongs to the technical field of underground processing. The method comprises the following steps of: separating common noise of a noise energy distribution matrix of a sensor and separating specific noise; carrying out adaptive noise set empirical mode decomposition on the separated signals, carrying out variational mode decomposition on residual signals in the signals, and screening effective modes through a kurtosis-entropy joint criterion to obtain signals with effective characteristic components reserved; high-frequency fluctuation of the signal is punished through total variation regularization, an improved alternating direction multiplier method algorithm is used for solving, short-time Fourier transform is carried out on the solved signal, low-frequency and high-frequency features are extracted through a multi-scale convolutional network, and a final denoised signal is output through gating weight fusion. According to the method, the signal denoising precision in a complex noise environment is remarkably improved, the processing time is shortened, and the resource consumption is reduced.

The invention discloses a laboratory detection data processing method and system based on a machine learning technology, and relates to the technical field of data processing. According to the method, the signal is decomposed through discrete wavelet transform, the noise standard deviation is calculated based on the median of the highest frequency coefficient, the high signal-to-noise ratio data is reconstructed through the dynamic threshold and the soft threshold function, and the signal quality is improved; a sliding window is used for extracting time sequence signal statistics and FFT frequency domain features, spatial features are extracted in combination with an SIFT algorithm, high-correlation features are reserved through mutual information screening, and redundancy is reduced; constructing a graph convolutional network anomaly detection model and an XGBoost-LightGBM weighted regression model, and eliminating pollution data through an anomaly probability threshold to obtain a normal regression predicted value; predicting a compensation amount according to the environmental parameters by using an LSTM network, and obtaining calibration laboratory data according to the normal regression predicted value and the predicted compensation amount; according to the method, the problems of noise sensitivity, feature splitting, model isolation and environment drifting of multi-source data are solved, and the laboratory analysis precision and robustness are remarkably improved.

The invention relates to a transformer multi-physics field coupling numerical simulation method and device and a storage medium. The method comprises the following steps: performing coupling calculation on a target working condition by using an electromagnetic, heat, flow and force coupling calculation model of a transformer to obtain transformer loss data and temperature rise data; mapping the temperature rise data to a transformer electromagnetic and force coupling calculation model by using a grid node data mapping method to serve as an initial value, and performing coupling calculation on a target working condition to obtain dynamic data of an internal structural member of the transformer; performing fast Fourier transform on the dynamic data in a preset frequency range to obtain dynamic data under each frequency; and mapping the temperature rise data and the dynamic data under each frequency to a transformer sound and force coupling calculation model as initial values, and carrying out coupling calculation on a target working condition to obtain transformer oil tank vibration and external noise data. Compared with the prior art, the method has the advantages that the calculation amount is effectively reduced, and the accuracy of predicting the internal state of the transformer is improved.

The invention relates to the technical field of laser spectrum detection, in particular to an LIBS (Laser-induced Breakdown Spectroscopy) spectrum noise reduction method, system and equipment based on adaptive threshold wavelet transform and a storage medium. Acquiring an original spectral signal of the laser-induced breakdown spectroscopy; performing five-layer multi-layer wavelet decomposition on the original spectral signal by adopting a db4 wavelet basis function to obtain a high-frequency coefficient and a low-frequency coefficient of each layer; calculating a noise intensity standard deviation based on the detail coefficient of the highest decomposition layer; dynamically determining the optimal value of the regulation factor through a double-layer optimization strategy combining a grid search method and a golden section iterative optimization method; constructing an adaptive threshold value based on the noise intensity standard deviation and the adjustment factor; carrying out threshold value processing on the high-frequency coefficient by adopting a self-adaptive threshold value; and performing wavelet reconstruction on the processed high-frequency coefficient and low-frequency coefficient, and outputting a denoised spectral signal. While the LIBS spectral signal-to-noise ratio is remarkably improved, the spectral feature form is completely reserved, and reliable technical support is provided for laser-induced breakdown spectroscopy detection in a complex industrial environment.

The invention belongs to the technical field of artificial intelligence, and relates to an artificial intelligence-based electric energy quality disturbance identification and positioning method, which comprises the steps of constructing an electric energy quality disturbance signal data set, performing segmented preprocessing on electric energy quality disturbance voltage data, enhancing time-frequency joint features and encoding disturbance sensitive areas. And constructing a deep learning model for power quality disturbance identification and positioning, and identifying and positioning the power quality disturbance. According to the invention, through adaptive denoising processing, boundary detection and multi-resolution time-frequency feature extraction, the identification precision and positioning precision of power quality disturbance are significantly improved; self-adaptive wavelet denoising and dynamic segmentation are combined, noise interference is effectively suppressed, and the edge characteristics of voltage sudden change points are kept; according to the dual-task sharing network, disturbance identification and positioning tasks are cooperatively optimized, so that the network can consider disturbance classification and time positioning at the same time; and through Bayesian reasoning, the system can output confidence estimation, provides credibility quantification of identification and positioning results, and effectively improves the reliability of the system.

The invention relates to the technical field of noise filtering, in particular to a communication line noise filtering system and method, and the system comprises a temperature coupling module, a voltage response module, a parameter joint debugging module and a link checking module. According to the method, inductance and capacitance node temperature data are collected and input into a piecewise linear regression model, the temperature-to-resonant frequency deviation gradient is calculated, a noise suppression bandwidth threshold value is dynamically matched in combination with a reference table, the frequency stability under temperature fluctuation is improved, Kalman filtering is used for processing voltage sampling, and a fluctuation predicted value is generated. According to the method, interval indexes are extracted based on a parameter mapping table, instant inductance and capacitance parameters are generated through least square fitting, voltage transient fluctuation self-adaption is achieved, noise power spectrum density is calculated through fast Fourier transform, a 3dB attenuation point is compared with a threshold value difference value to trigger parameter rollback, closed-loop control is formed, it is ensured that suppression bandwidth is continuously matched with a noise frequency band, and noise suppression is achieved. The bandwidth offset is reduced to be within 1.2%, and the optimization response time is shortened to 200 ms level.

The invention relates to the technical field of intelligent sensor application, and discloses a water meter flow real-time correction method and system based on an intelligent sensor, and the system comprises a flow signal collection module, a signal feature decomposition module, a frequency domain feature fusion module, an error compensation decision module and a correction instruction execution module. The method comprises the following steps: acquiring multi-dimensional vibration waveform data of a water meter pipeline through a multi-band sensor, performing time-frequency domain decomposition and energy spectrum feature extraction to obtain a local frequency domain feature tensor set, and performing dynamic noise suppression and band correlation optimization to obtain a local optimization feature tensor set; and performing global feature reconstruction based on frequency domain energy distribution to generate an error compensation decision feature map, and finally generating a real-time correction control signal. Through multi-dimensional signal acquisition and intelligent feature processing, the precision and dynamic adaptability of flow metering in a complex pipe network environment are improved, and the method is suitable for high-precision flow correction of an intelligent water meter and has the advantages of being high in anti-interference capacity, accurate in compensation and the like.

The invention relates to the technical field of information transmission, in particular to a multi-band processing system for electronic communication, which comprises a signal separation module, a frequency deviation compensation module, an amplitude correction module, a phase correction module and a dynamic feedback adjustment module. According to the invention, a plurality of frequency bands are extracted to execute band-pass filtering to separate signal components, short-time power spectral density is calculated to describe energy distribution, normalization coefficient operation is carried out to enhance frequency domain balance, high-energy component screening is carried out to reduce noise interference, signal effectiveness is improved, and frequency offset compensation calculates a change rate based on adjacent phase data and eliminates abnormal fluctuation. A time domain coefficient is adjusted to correct a frequency offset error, amplitude correction depends on short-time Fourier transform to screen the maximum amplitude to optimize signal consistency, phase adjustment fits a group delay error curve to optimize compensation parameters, dynamic feedback optimizes a strategy according to a signal-to-noise ratio to enhance compensation precision, and signal transmission quality and stability are improved.

The invention relates to the technical field of signal analysis, and discloses a primary frequency modulation large disturbance signal peak value analysis method, which comprises the following steps: collecting a primary frequency modulation large disturbance signal and carrying out filtering processing of multi-band noise separation; performing multi-scale decomposition on the filtered disturbance signal, extracting a dynamic entropy of a multi-scale signal sequence, and performing dynamic entropy denoising on the disturbance signal to obtain a denoised disturbance signal; and calculating a potential peak position of the denoised disturbance signal, and carrying out multi-condition verification on the potential peak position to obtain peak information for frequency modulation disturbance analysis. Based on spectrum energy distribution, specific wavelet bases are selected for different noise frequency bands for noise reduction, the wavelet noise reduction performance is improved, the energy distribution complexity and the structure complexity of the disturbance signals on different scales are extracted, the dynamic characteristics of the disturbance signals are captured more comprehensively, the noise reduction precision is improved, and the noise reduction efficiency is improved. And the accuracy of wind power integration fault identification and equipment life monitoring in the peak value analysis process is improved.

The invention discloses a self-adaptive test parameter optimization method, and relates to the technical field of parameter optimization, and the method comprises the steps: collecting an original sensing signal in a mechanical test system, and carrying out the preprocessing of the original sensing signal; based on the preprocessed data set, constructing a four-dimensional space-time tensor, and executing improved parallel factor tensor decomposition to obtain a decoupled core factor and space-time feature component matrix; performing cross-modal alignment on the decoupled core factor and the time-space feature component matrix through a wear feature channel and an acoustic emission feature channel to obtain a fused cross-modal feature vector; performing crack growth rate prediction on the fused cross-modal feature vectors to obtain a crack risk level, and adjusting a strategy through dynamic parameters to obtain an optimized parameter set; the problem of signal noise and time mismatch is solved through multi-mode signal preprocessing, and high signal-to-noise ratio input is provided for subsequent analysis in combination with wavelet denoising, space-time alignment and double-domain feature extraction.

The invention relates to the field of image data processing, and discloses a texture preserving type image denoising and enhancing method based on a generative adversarial network, which comprises the following steps: acquiring an original image signal, and calculating low-frequency sub-band data and high-frequency sub-band data by using discrete wavelet transform; calculating the gradient magnitude of the low-frequency sub-band data to generate a structural significance gradient map; establishing a reverse mapping relation based on the structure saliency gradient map, and generating a spatial self-adaptive dynamic gating threshold; performing statistical gating on the high-frequency sub-band data by using the dynamic gating threshold to generate a high-pass gain coefficient and a low-pass suppression coefficient; according to the method, cross-band modulation logic of the structure flow to the texture flow is established, so that the problem that weak texture signals are easy to lose under non-uniform illumination is solved, and non-structured noise filtering and structured micro texture restoration are realized on the premise of not depending on semantic tags.

The invention provides a joint denoising method and system based on adaptive large neighborhood search and modal decomposition, and belongs to the technical field of signal processing and nondestructive detection.The method comprises the steps that an ultrasonic signal and a vibration signal of a detected insulator are synchronously collected and preprocessed; dynamically estimating the noise level based on the preprocessed ultrasonic signal power spectral density, and optimizing decomposition parameters by adopting an adaptive large neighborhood search algorithm; on the basis of the optimized decomposition parameters, wavelet packet decomposition and ensemble empirical mode decomposition are executed in parallel, and effective intrinsic mode function components are screened through cross-correlation verification; extracting the resonance frequency of the preprocessed vibration signal, performing target frequency band weighted enhancement on the low-frequency sub-band, and dynamically adjusting the threshold parameter of the high-frequency sub-band and the low-frequency sub-band according to the resonance frequency; and generating a preliminary de-noised signal from the fused signal, performing affine projection algorithm filtering and multi-modal cross validation, and outputting the verified ultrasonic signal as a final de-noising result.

The invention provides an adaptive welding seam detection and three-dimensional reconstruction method based on deep learning and binocular vision. The adaptive welding seam detection and three-dimensional reconstruction method comprises the steps of S1, collecting samples and making a training data set; s2, the picture of the sample to be welded is processed, a feature region is recognized, the image quality of the region to be welded is analyzed and evaluated through wavelet transform and local variance, and the noise level and the contrast ratio are calculated; s3, dynamically generating edge detection parameters and model fitting parameters according to the image quality; s4, using an edge detection algorithm to extract edge point cloud of the welding seam area; s5, performing RANSAC linear fitting, weighted least square fitting and polynomial curve fitting on the edge point cloud in parallel; s6, selecting an optimal fitting result based on an image quality adaptive dynamic scoring model; and S7, carrying out three-dimensional coordinate conversion in combination with the three-dimensional matching model IGEV-Stereo, and outputting a final welding seam three-dimensional coordinate. According to the invention, automatic detection of the position and size of the welding seam can be efficiently and accurately realized.

The invention belongs to the technical field of diving state monitoring, and relates to a diver state monitoring method based on biofeedback, and the method comprises the steps: aligning a sign signal with environment data through a timestamp, extracting motion features through acceleration data, constructing a motion noise model, and removing signal artifacts in real time; in the aspect of EEG signal purification, a db4 wavelet basis is adopted for carrying out wavelet packet decomposition, water pressure related frequency band characteristics are extracted, low-frequency drift is removed, meanwhile, heart rate signals are analyzed through nonlinear dynamics, a Lyapunov index is extracted to evaluate HRV changes, breathing and EEG # imgabs0 # wave coupling are analyzed through wavelet transformation, and the risk evaluation precision is further enhanced; in the aspect of environment compensation, through compensation of a depth and SpO2 relational expression and water temperature on respiratory frequency, in combination with SLTM network extraction time sequence features and a deep causal convolutional layer, risk trend prediction is performed, a diver risk level and a prediction curve are output in real time, a comprehensive risk index is finally calculated, the health state and potential risk of the diver are comprehensively evaluated, and safety is improved.

The invention discloses a loss tuning method of a power transformer, which is applied to a transformer body sleeved with a winding and comprises the following steps: applying scanning current excitation containing fundamental waves and harmonic waves to the winding, synchronously acquiring a body vibration signal and converting the body vibration signal into a frequency spectrum; extracting a formant from the frequency spectrum, matching the formant with a theoretical electromagnetic force wave and a structure inherent frequency library, and identifying a coupling formant to be optimized; aiming at each formant, installing a vibration exciter in a corresponding area, sending out an anti-phase periodic pulse force, and dynamically and finely adjusting a pulse force parameter by monitoring a vibration response in real time and taking equivalent mechanical impedance minimization as a target; when the optimal damping state is achieved, the vibration exciter output rod is locked, and static pre-tightening force is formed; and after all formants are adjusted and optimized in sequence and the prestress is locked, final pressing and fixing of the transformer body are completed in the state that the pretightening force is kept. According to the invention, the dynamic loss source of the individual transformer can be actively inhibited and cured before assembly and curing, the operation loss and noise are effectively reduced, and the structural stability is improved.

The invention discloses a phytoplankton microscopic image classification model construction method, and belongs to the field of intelligent detection. For the problems of unstable microscopic image quality and small difference between phytoplankton classes, a double-branch ResNeXtViT model is provided: a ResNeXt branch enhances the texture feature extraction capability through a wavelet convolution attention module, and improves the adaptability to irregular shapes in combination with an adaptive convolution module; a Transform branch is used for capturing global context information; the feature fusion module integrates local details and a global structure. A combined data enhancement strategy including noise addition and color migration is adopted, and a joint loss function is designed to optimize intra-class compactness and local discrimination capability. Experiments show that the accuracy of the method on a Chaohu lake phytoplankton data set reaches 94.11%, which is superior to that of ResNet, ViT and other models, and the method can be deployed to an intelligent analyzer to realize rapid water quality monitoring.

According to the cooling tower early fault early warning method based on vibration state monitoring, vibration signals and working condition labels of key parts of the cooling tower are synchronously collected through multiple channels, and data quality is improved through preprocessing operation such as band-pass filtering and normalization; time-frequency features are extracted in a multi-scale mode through self-adaptive variational mode decomposition and wavelet packet transformation, signal complexity is quantized through energy entropy, and weak fault detection capacity is enhanced; the obtained features are input into a deep belief network after being subjected to principal component analysis dimensionality reduction, and automatic classification and recognition of the equipment operation state are achieved; dynamic early warning grade adaptation is carried out according to an identification result in combination with a working condition label, the environmental adaptability and stability of early warning are effectively improved, the method further has the functions of early warning sample recording and periodic model iterative optimization, and the fault identification precision and robustness in a complex noise environment are remarkably improved.

The invention provides an under-mine video stream dynamic denoising method based on multi-modal fusion, which comprises the following steps: constructing a time sequence synchronous fusion mechanism of visible light, infrared and laser radar data, and realizing time-space alignment of multi-source heterogeneous data; a dynamic noise model is established by introducing a fractional calculus optical flow field concept and combining a Gaussian mixture model, so that a dynamic noise region is accurately identified; an improved self-adaptive wavelet threshold function is constructed, a function threshold parameter can be linked with a dust concentration sensor in real time, and the de-noising intensity is dynamically adjusted according to the actual dust concentration; designing a dual-path feature enhancement neural network to effectively separate and enhance structural features and texture features in the video image; a cascaded detection decision system is created, a lightweight network is used as a primary detector, a high-confidence detection result is directly output, and a low-confidence detection result is input into a Transform correction module for secondary reasoning. According to the invention, dynamic denoising, feature enhancement and target intelligent monitoring of the video stream under the mine can be realized.

The invention discloses an improved filtering and local direction adaptive rust recognition method for a hull surface rust image, which comprises the following steps of: 1, performing gray processing on the hull surface rust image to obtain a gray image; 2, performing global denoising on the image after gray scale by adopting a wavelet transform and total variation fusion denoising algorithm to obtain a globally denoised image; 3, performing local denoising on the globally denoised image by adopting a self-adaptive Gaussian kernel denoising algorithm to obtain a final denoised image; 4, carrying out the edge extraction of the final denoised image through a direction self-adaptive edge detection algorithm, and obtaining a rust image; according to the method, the improved wavelet transform and total variation fusion denoising algorithm is combined, efficient noise removal is achieved, the size and the standard deviation of the filtering kernel are adjusted in a self-adaptive mode, noise of a local area is effectively processed, detail loss is avoided, and the image quality is improved.

The invention discloses a time series data anti-noise anomaly detection method and system based on dynamic decomposition, and the method comprises the steps: obtaining a to-be-detected time series, inputting an anomaly detection model, and carrying out the dynamic decomposition of the time series; based on a channel-time mixed attention mechanism, obtaining channel feature representation of a trend component and time feature representation of a seasonal component; and obtaining a reconstruction sequence according to the channel feature representation and the time feature representation, calculating a difference value of each corresponding time step in the reconstruction sequence and the time sequence, and comparing and marking the difference value with an abnormal threshold value to obtain a marked abnormal detection sequence so as to complete sequence abnormal detection. According to the method, the time sequence needing to be detected is subjected to dynamic trend seasonal decomposition through learnable one-dimensional convolution and discrete Fourier transform of the anomaly detection model, the potential time pattern of the time sequence can be effectively extracted, the adaptability of the model to complex trends and periodic changes is improved, and the accuracy of abnormal data recognition is improved.