250 results about "Adaptive wavelet" patented technology

The invention relates to the technical field of wind turbine generator fault monitoring. The invention provides a wind turbine generator fault monitoring method and system. The method comprises the following steps: synchronously acquiring gearbox and environment temperature and humidity data, and generating a time-frequency energy fusion matrix through adaptive wavelet packet transformation; adopting mutual information entropy weighted improved variational mode decomposition to screen out an intrinsic mode component set related to a fault mode; constructing a space-time double-flow residual network based on the intrinsic mode component set, and fusing two branch outputs of the space-time double-flow residual network through a dynamic feature gating mechanism to obtain a multi-dimensional feature vector; and inputting a multi-dimensional feature vector obtained by fusion into a lightweight fault classifier, and outputting a real-time fault probability and a component health degree evaluation index based on a sliding window mechanism. The problems of low efficiency, high false alarm rate, missing detection of early faults, reduction of prediction precision, incapability of mining multivariable coupling relations, need of massive annotation data, and high delay caused by insufficient edge side computing power existing in an existing wind turbine generator fault monitoring mode are solved.

The invention relates to a supporting structure stress state monitoring method based on artificial intelligence, and belongs to the technical field of artificial intelligence and data processing. The method comprises the following steps: acquiring and marking strain data of a supporting structure; after abnormal values are removed, normalizing the multi-sensor data to generate a normalized strain sequence; a state monitoring model is constructed, a deep time sequence neural network architecture is adopted, and the state monitoring model comprises an input layer, a self-adaptive wavelet attention feature mapping layer, a time domain gating convolution module, a global maximum pooling layer, a dynamic feature importance reweighting layer and a full-connection classification layer; inputting a normalized data training model; optimizing a loss function through a quantile interval adaptive learning rate and a momentum updating strategy; after real-time monitoring data is processed, inputting the data into the training model according to time window slices, outputting four types of probabilities, and taking the maximum value as a prediction state; and if a plurality of continuous windows are early-warning and dangerous, triggering the terminal to give an alarm. The accuracy of monitoring the stress state of the supporting structure can be improved.

The invention relates to the technical field of nondestructive testing, and discloses an ultrasonic nondestructive testing method and system for a welding seam of a steel structure. The method comprises the following steps: adopting a self-adaptive coupling ultrasonic probe to carry out signal acquisition on the surface of a welding seam of a steel structure to obtain original ultrasonic signal data; performing multispectral adaptive wavelet transform processing on the original ultrasonic signal data to obtain an enhanced ultrasonic signal; performing blind source separation on the enhanced ultrasonic signal to obtain a separated weld defect signal; performing triangular constraint average interpolation and adaptive extended Kalman filtering fusion processing on the separated weld defect signals to obtain filtered weld defect feature data; and performing regularization Gaussian field weld contour reconstruction and defect positioning processing based on the filtered weld defect feature data to obtain a weld contour model and defect position data. According to the invention, the stability and reliability of weld defect detection are improved, and high-precision detection of weld internal defects is realized.

The invention discloses an acoustic emission intelligent detection method and system for hydrogen-induced damage of a high-pressure hydrogen system, and the method comprises the steps: collecting a system operation signal through an acoustic emission sensor, carrying out the combined preprocessing of variational mode decomposition and adaptive wavelet threshold noise reduction, restraining noise, constructing a lightweight MobileNet-TCN network, carrying out the deep feature extraction, and carrying out the detection of the hydrogen-induced damage of the high-pressure hydrogen system. GRU and a three-dimensional point cloud technology are fused to realize submillimeter-level positioning of an acoustic emission source, a big data damage case library is associated based on an acoustic emission parameter accumulation model, dynamic assessment and early warning of damage risks are realized, multi-physics field monitoring data are combined, damage classification is optimized through a GCNs (Graph Convolutional Networks), and the above processes are systematically integrated. And full-chain intelligent processing from signal acquisition to risk early warning is completed. The scheme has the advantages of strong anti-interference capability, submillimeter positioning precision, high edge end reasoning efficiency, high damage classification accuracy and dynamic early warning capability, and is suitable for safety monitoring of hydrogen energy storage and transportation equipment.

The invention discloses a space-frequency domain graph learning-based adaptive deep fake face detection method and system, and the method comprises the steps: randomly extracting an image frame from a video, intercepting a face image, adjusting the feature dimension of the face image, and transmitting the face image to a depth adaptive wavelet module and a normalized residual homomorphic composition neural network module; a depth adaptive wavelet module extracts frequency features of the face image; a normalized residual homograph neural network module extracts spatial domain features of the face image; performing weighted fusion on the frequency domain features and the spatial domain features by using a self-adaptive feature fusion module based on gated convolution, realizing class attention guidance by using the gated convolution, dynamically adjusting the channel of a feature map and the weight of a spatial dimension, and finally obtaining fusion features; and performing classification according to the fusion features by using a classifier. According to the method, the extraction capability of forged detail clues is enhanced, and the detection precision and stability of the model are remarkably improved.

The invention relates to the field of equipment diagnosis, and particularly discloses an equipment predictive maintenance method based on deep learning, and the method comprises the steps: carrying out the time domain amplitude normalization and frequency domain weighted normalization of training data, and carrying out the dual-channel feature fusion, so as to obtain a fusion feature vector; performing feature extraction on the fused feature vector by using multiple groups of self-adaptive wavelet kernels to obtain a self-adaptive time-frequency feature vector; constructing a weight population through statistical characteristics, and screening the optimal initial weight of the initial diagnosis model; carrying out multi-scale depth feature calculation, time-frequency domain attention feature fusion, fault prototype comparative learning and a pre-constructed total loss function on the adaptive time-frequency feature vector, and carrying out iterative updating on the initial diagnosis model to obtain a diagnosis model; and the equipment is diagnosed through the diagnosis model. Multi-scale feature fusion and a double-path attention mechanism can cooperatively capture short-time impact and a long-period mode, and the limitation of a traditional method in diversified fault scenes is overcome.

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 discloses a digital twinning system and a construction method thereof, and relates to the technical field of digital twinning, and the construction method comprises the steps: carrying out the classified collection of multi-source heterogeneous data of a digital twinning entity, carrying out the filtering and denoising through a self-adaptive wavelet threshold, mapping the data into a multi-dimensional feature vector, screening key features, and outputting the key feature data; the method comprises the following steps: establishing a physical mechanism layer based on a general physical rule, defining core physical parameters and a constraint equation, initializing a particle swarm to establish a data driving layer, constructing an LSTM time sequence prediction module, executing particle filter state calibration, and selecting a model to configure a communication protocol to establish a virtual-real interaction layer, so as to realize state mapping and instruction feedback of a digital twin entity and a virtual model; and calculating virtual and real state deviation in real time and performing attribution diagnosis, adjusting model parameters or key features according to a deviation source, generating an optimization parameter combination based on a long time sequence prediction result and performing simulation verification in a virtual environment, and controlling entity parameter adjustment through an instruction feedback channel so as to realize predictive optimization.

The invention belongs to the technical field of signal acquisition and processing, and discloses a low-noise biopotential signal acquisition system and a low-noise biopotential signal processing method. An original bioelectricity signal of a human body is collected through an electrode array, and analog-to-digital conversion is carried out after the signal is processed by a multi-stage self-adaptive filtering and amplifying circuit. A digital signal is sequentially subjected to adaptive wavelet transform denoising and empirical mode decomposition to obtain a multi-level intrinsic mode function set, an intrinsic mode function of a characteristic frequency band is extracted from the multi-level intrinsic mode function set, and the signal is reconstructed. Introducing a dual noise reduction mechanism combining adaptive wavelet transform and empirical mode decomposition; a signal quality real-time evaluation system is established, and system parameters including the gain of a variable gain amplifier and the bandwidth of a dynamic band-pass filter are dynamically adjusted through closed-loop feedback. The self-adaptive optimization of the whole biopotential signal acquisition process is realized, various interferences are effectively inhibited, the signal characteristics are reserved, and the signal-to-noise ratio is 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.

A system and method are disclosed for low-light image enhancement using hierarchical adaptive wavelet decomposition with cross-scale feature fusion. The system analyzes a raw input image to determine image characteristics and preprocessing parameters. A hierarchical adaptive wavelet decomposition process creates a variable-depth decomposition tree comprising frequency domain nodes, with decomposition depth determined by local image complexity. Cross-scale feature fusion implements attention mechanisms between nodes at different decomposition levels, enabling bidirectional information flow across scales. A dynamic network pool allocates specialized neural networks to process nodes based on their frequency characteristics, with weight sharing between similar nodes for efficiency. An adaptive reconstruction engine traverses the decomposition tree using learned filters and multi-scale residual learning to produce an enhanced image. The hierarchical approach enables superior low-light image enhancement by allocating computational resources based on content complexity, achieving better quality than fixed decomposition methods while maintaining compatibility with existing image signal processing pipelines.

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 a lightweight super-resolution system and method of an adaptive wavelet attention network, and belongs to the technical field of image processing. The system is composed of a multistage wavelet attention module, a dynamic convolution kernel generation unit, a convolutional neural network and an output unit. The method comprises the following steps of: extracting features of a low-resolution image and performing Haar wavelet decomposition; calculating a cross-scale attention weight on each high-frequency sub-band and carrying out weighted fusion to highlight details; adaptively generating a dynamic convolution kernel of a Haar wavelet basis kernel weighted combination based on the fused features, and performing directional convolution enhancement on the features; high-resolution image reconstruction is realized through a lightweight residual network and pixel rearrangement; and during training, pixel domain mean square error and wavelet coefficient compensation loss joint optimization is adopted. The parameter quantity of the system model is smaller than 450KB, a 1080p video super-resolution task can be processed on mobile equipment in real time, and the texture recovery performance is improved by about 1.2 dB compared with that of an existing lightweight model.

The invention discloses a multi-mode short-term photovoltaic power prediction method and device based on a satellite cloud atlas, and the method comprises the steps: obtaining real-time meteorological data, historical photovoltaic power data and cloud cluster image data, and carrying out the preprocessing of the data; a GPAformer model is established, and the meteorological data after noise reduction are predicted; the method comprises the following steps: establishing an SA-Convlstm model, capturing a cloud cluster movement track, extracting time change characteristics, predicting cloud cluster image data, introducing a cloud shielding model, and correcting the prediction deviation of SA-ConvLSTM under the condition that a cloud layer is dense or changes rapidly; establishing a photovoltaic power KAN-COGCN prediction model, and performing photovoltaic power prediction by taking the meteorological factors, the cloud cluster motion time sequence characteristics and historical photovoltaic power data obtained by prediction of the GPAformer and the SA-ConvLSTM model as input; using an improved alpha evolutionary algorithm IAE to optimize hyper-parameters of the three models; and performing error correction on a prediction result by establishing an adaptive wavelet RBF neural network to obtain a final prediction result. According to the invention, the precision and effectiveness of photovoltaic power prediction can be improved.

The invention discloses a transformer sound signal adaptive wavelet optimization and feature extraction method and system, and the method comprises the steps: calling a Pywt wavelet analysis library, decomposing an original signal according to a decomposition layer number J, and obtaining a multi-layer detail signal; for each layer of detail signals, the following steps are executed: introducing an M estimator to improve a noise variance calculation model, and calculating a standard deviation and a unified monitoring threshold value; constructing a dynamic threshold value based on a denoising signal approximation error minimization criterion; designing a correction factor; correcting the wavelet coefficient of each layer of detail signal; reconstructing a pure signal by using an inverse decomposition method; dividing the pure signal into a plurality of short-time signals; extracting an MFCC feature vector of each short-time signal; weighting and screening MFCC feature vectors by adopting a support vector machine recursive feature elimination method; and compressing the dimension of the feature vector in combination with a principal component analysis algorithm to generate a final feature matrix. According to the method, the problems of contradiction between noise suppression and signal fidelity and low recognition rate caused by high-dimensional feature redundancy in a traditional method are solved.

The invention provides a partial discharge on-line monitoring method and system based on multi-modal fusion and adaptive noise reduction, and the method comprises the steps: employing an ultrahigh frequency UHF sensor, a miniature ultrasonic sensor, and a miniature detector for gas dissolved in oil, which are disposed on a transformer; synchronously acquiring electric signals, sound signals and characteristic gas concentration data in oil generated in the partial discharge process; performing format unification, abnormal value elimination and time alignment processing on the electric signal, the sound signal and the gas concentration data in an edge calculation unit to obtain aligned multi-source original data; performing adaptive wavelet noise reduction processing on the electric signal to obtain a de-noised UHF signal; respectively extracting time domain, frequency domain and chemical features from the de-noised UHF signal, the aligned sound signal and the gas concentration data to form a multi-dimensional feature vector; and inputting the multi-dimensional feature vector into a pre-trained defect traceability model, and outputting a partial discharge defect type, severity level and development trend prediction result.

The invention discloses a phase modifier unit state monitoring and load control method and system for reactive power compensation of a power grid, belongs to the technical field of automatic control of power systems, and mainly solves the problem that service life loss is aggravated or a control target is single due to the fact that the health state of equipment is not fully considered when an existing phase modifier unit responds to reactive power requirements of the power grid. The method comprises the following steps of: calculating reactive vacancy and extracting vibration frequency band energy characteristics by acquiring voltage and frequency of a power grid as well as vibration, temperature and exciting current signals of a unit in real time, and constructing a health state index HSI in combination with self-adaptive wavelet basis selection; taking minimum equipment life loss and optimal power grid voltage stability as multiple objectives, dynamically generating a load control instruction by adopting a particle swarm algorithm, realizing weight adaptive adjustment based on HSI, and introducing an online life loss model as a constraint condition; finally, reactive power output is adjusted through an excitation system, and cooperative control of power grid stability and equipment life extension is achieved.

The invention discloses a distribution transformer state evaluation method, system and device and a storage medium, and the method comprises the steps: obtaining current and voltage long sequence data through high-frequency sampling, and forming a high-quality data set through expert labeling; signal preprocessing is realized by adopting dual-channel adaptive wavelet packet denoising and cross-correlation sinchinger interpolation phase correction; constructing an asymmetric convolution pyramid to extract current high-frequency and voltage low-frequency multi-scale features, and performing cross-modal fusion by using compressed multi-head attention and a channel-space double-gating mechanism; dynamic random depth regularization is introduced, and a time domain loss function in a fault sensitive period is focused, so that high-accuracy and low-delay real-time state evaluation is realized; the capacity of capturing and recognizing early weak fault features of the distribution transformer in a complex operation environment is remarkably enhanced, and therefore more sensitive and more accurate state early warning is achieved.

The invention discloses an airborne radar identification method based on multi-scale signal decomposition and depth feature fusion, belongs to the technical field of radar target identification, is used for unmanned aerial vehicle radar target identification, and comprises the following steps: decomposing an original dynamic radar cross section signal into a plurality of intrinsic mode components by adopting an adaptive noise complete set empirical mode decomposition method; self-adaptive wavelet soft threshold denoising is carried out on all the components in a unified mode so as to improve the signal-to-noise ratio and suppress background interference; effective components are screened out by calculating correlation coefficients of all intrinsic mode components and original signals, a radar cross section signal sequence after noise reduction is reconstructed, and multilevel feature fusion is carried out; and inputting the feature vector into a stacked auto-encoder deep neural network classification model, and introducing a particle swarm optimization algorithm to optimize the classification model. According to the method, the signal-to-noise ratio of the signal and the recognition accuracy of the classification model in a low signal-to-noise ratio environment are remarkably improved, and high-precision recognition and classification of different types of radar cross section targets are realized.

The invention relates to the technical field of atmospheric remote sensing, in particular to a satellite-borne laser radar-based multi-feature attention cloud top height estimation algorithm, which comprises the following steps of: full-waveform laser radar data acquisition and preprocessing: acquiring an original echo signal through a satellite-borne full-waveform laser radar system; performing time domain alignment, noise suppression and saturated signal elimination on the signals; a median filtering and wavelet threshold denoising algorithm is combined to eliminate high-frequency noise, and background radiation interference is eliminated through dynamic baseline correction; according to the method, a multi-stage noise reduction strategy combining sliding window median filtering and self-adaptive wavelet denoising is introduced in the waveform feature extraction stage, pulse interference and signal oscillation are effectively restrained, meanwhile, through a terrain matching error compensation mechanism, the detection accuracy is improved, and the detection accuracy is improved. The physical consistency of the elevation residual error under the real terrain background is enhanced, and the authenticity and the checkability of the point elevation return are improved.

The invention relates to the technical field of fault diagnosis, and discloses a cable fault diagnosis method, system and terminal, and the method comprises the steps: applying a broadband excitation signal to a target cable, so as to excite the target cable to generate a first response signal; extracting multi-dimensional time-frequency features of the first response signal after adaptive wavelet threshold denoising, and matching a fault template of a preset cable reference model according to energy distribution features in the multi-dimensional time-frequency features; determining a fault area in a fault template corresponding to the target cable according to the propagation time and the propagation speed of the first response signal; applying the narrowband excitation signal to a fault area of the target cable to obtain an impedance spectrum; the resonant frequency point based on the impedance frequency spectrum and the corresponding impedance amplitude value are compared with a typical fault mode in a fault feature database, and the fault type and the fault position of the target cable are obtained; generating a diagnosis report of the target cable according to the fault type and the fault position; the cable fault diagnosis accuracy can be improved.

The invention relates to a GIS partial discharge optical detection method and device, and a medium. The method comprises the following steps: collecting a GIS partial discharge optical signal through a light guide rod, and carrying out the preprocessing of the GIS partial discharge optical signal; self-adaptive wavelet packet decomposition is carried out on the preprocessed signals, noise signals are separated through a self-adaptive threshold value adjustment algorithm, multi-scale time-frequency characteristics are extracted from the signals after noise separation, and according to the self-adaptive wavelet packet decomposition, a wavelet basis function is selected in a self-adaptive mode according to statistical indexes of the extracted signals so as to carry out wavelet packet decomposition; and constructing a lightweight convolutional neural network model, classifying the extracted multi-scale time-frequency features, and identifying a partial discharge signal. Compared with the prior art, the method has the advantages that the anti-interference capability is high, weak partial discharge signals can be effectively extracted, and the real-time requirement is met.

The invention discloses an infrared anti-unmanned aerial vehicle detection method based on an improved YOLO framework. The infrared anti-unmanned aerial vehicle detection method comprises the following steps: S1, preprocessing an infrared image and calculating energy distribution; s2, embedding the backbone network into an improved convolution module and adaptive wavelet basis convolution, and optimizing feature extraction in stages; s3, reinforcing local details and global contours of the feature maps by a multi-granularity pooling mechanism; s4, realizing cross-layer feature fusion through a feature aggregation propagation mechanism, and generating a comprehensive multi-scale feature map; s5, a lightweight detection channel is specially arranged to enhance the small target detection capability, and other standard detection heads are responsible for classified positioning of medium and large targets; and S6, outputting a final prediction result through a context sensing weighted frame fusion algorithm. Through adaptive wavelet feature enhancement, multi-granularity feature optimization, a lightweight detection head and an intelligent frame fusion technology, the problems of small target leak detection, background interference and multi-scale detection real-time performance in an infrared scene are solved, and the all-weather detection capability and the positioning precision are remarkably improved.

The invention relates to a micro-grid frequency control system with hybrid energy storage. The system comprises an input power decomposition module for performing adaptive wavelet packet decomposition on wind power generation output power to obtain an optimal low-frequency component and an optimal high-frequency component; the initial power distribution module is used for decomposing the high-frequency component into a sub-high-frequency component and an ultrahigh-frequency component and respectively distributing the sub-high-frequency component and the ultrahigh-frequency component to a storage battery and a super capacitor to obtain a primary power distribution instruction; the correction power distribution module is used for correcting the primary power distribution instruction through a Logistic regression function to obtain an optimized output instruction; and the power grid primary frequency modulation module inputs the frequency deviation and the frequency change rate of the micro-grid into a hybrid energy storage model based on virtual inertia control and virtual droop control to obtain a total frequency modulation output demand and generate a total frequency modulation output instruction. According to the system, through the adaptive wavelet packet decomposition and hybrid energy storage technology and in combination with virtual inertia and virtual droop control, the fluctuation of the distributed power supply can be effectively stabilized, and the frequency stability of the micro-grid is improved.

The invention relates to the cross technical field of geological disaster monitoring and signal processing, and provides a GNSS tunnel portal slope deformation time sequence noise reduction method, system and medium combining CEEMD and an adaptive wavelet packet threshold function, the method integrates the advantages of CEEMD and wavelet packet adaptation, processing of mixed entropy demarcation, adaptive threshold, dynamic smoothing and multi-path compensation is used, and the noise reduction efficiency is improved. Useful deformation features can be reserved while more comprehensive and finer noise reduction is performed, smooth compression can be performed in a large signal part through improved threshold function processing, excessive suppression can be avoided, smooth transition is performed in a small signal part through exponential attenuation, features of the large signal part are reserved, noise suppression is realized, signal details are reserved to the greatest extent, and the noise suppression effect is improved. The noise reduction requirement of tunnel portal side slope monitoring data is met, and particularly in side slope monitoring of a tunnel portal in a complex environment, the precision and reliability of GNSS deformation monitoring are improved.

The invention relates to a building structure health monitoring method and system based on machine learning, and the method specifically comprises the following steps: firstly, building a target building three-dimensional numerical model through finite element simulation, generating a simulation signal, injecting Gaussian white noise, and adjusting model parameters to form a data set with health category labels; performing data enhancement by combining adaptive wavelet denoising with dynamic normalization, and extracting and enhancing high-resolution time-frequency features through adaptive window short-time Fourier transform and adaptive frequency band enhancement; then, a neural network model fusing structure physical prior guidance and multi-scale space-time interaction is constructed, a feature matrix is modulated, fused and coded to obtain a refined feature vector, and damage state probability distribution output is achieved; and training is carried out by using a feature consistency and prediction smoothness regularization term constraint model, and finally, the trained model is deployed, so that building structure health state evaluation and safety early warning are realized, the monitoring accuracy and reliability are improved, and effective technical support is provided for building safety guarantee.

The invention discloses a vegetable disease incubation period detection method and system based on bimodal time sequence collaborative fusion, and the method comprises the steps: obtaining a leaf image through building an acquisition environment, constructing a training sample data set, and constructing a downy mildew incubation period spectral feature adaptive enhancement (AW-FPF) module; the method comprises the following steps: decomposing a hyperspectral signal into low-frequency and high-frequency components through spectrum time sequence adaptive wavelet decoupling, obtaining an enhanced feature tensor through spectrum multi-scale pathological feature frequency-time dual-path aggregation fusion frequency domain and time domain paths, and obtaining a first feature sequence through weighted screening by using a multi-head attention mechanism; a downy mildew incubation period prediction (HyChl-TFNet) model containing a hyperspectral branch, a chlorophyll fluorescence parameter branch, a feature fusion branch and a classifier is constructed, bimodal features are processed and fused to output a day number prediction result, accurate recognition of the downy mildew incubation period is achieved, the detection precision can be controlled to the day number level, and the detection accuracy is improved. And an accurate time basis is provided for early prevention and control of diseases.

The invention belongs to the technical field of wireless sensing and artificial intelligence, and discloses a high-robustness non-contact accurate electrocardiogram monitoring method based on a millimeter wave radar. Firstly, the distance and angle of a potential target are obtained through distance fast Fourier transform and digital beam forming technologies, and static background removal and thoracic cavity position detection are achieved in combination with mean filtering and a two-dimensional constant false alarm rate algorithm. And then a continuous phase is extracted by using a differential cross multiplication method, a two-step heartbeat-related phase extraction scheme is designed, body micro-motion and breathing interference are removed by adopting B-spline fitting and differential operation respectively, and a stable heartbeat-related phase signal is obtained. A heart rate-guided adaptive wavelet decomposition method is designed to obtain multiband features, and time-frequency joint features are extracted through a double-branch attention mechanism and a gating fusion part. Finally, the time-frequency joint features are input into an electrocardiosignal time domain reconstruction module based on a TransUNet architecture, high-quality reconstruction of electrocardiosignals is achieved, and the method has the advantages of being non-contact, continuous and convenient.

The invention discloses a vibration signal processing method based on self-adaptive wavelet packet and deep learning fusion, and belongs to the field of sewage plant equipment fault diagnosis. The method aims at solving the problems that traditional signal processing is poor in flexibility, the non-stationary signal processing capacity is weak, the deep learning data requirement is large, and the high-frequency weak feature capturing capacity is limited. According to the method, the high-frequency acceleration sensor is adopted, the vibration signals of the sewage plant equipment are accurately collected, the self-adaptive wavelet packet decomposition technology is applied, the primary function is dynamically selected, the number of decomposition layers is optimized, self-adaptive threshold noise reduction is achieved, and the signal processing quality is improved. Meanwhile, in combination with a one-dimensional convolutional neural network and a bidirectional LSTM model, local and global features of the signal are extracted respectively, and pre-processed data are formed through gating weighted fusion. According to the method, the signal-to-noise ratio and the weak fault detection rate are remarkably improved, feature redundancy and data requirements are reduced, the calculation efficiency and diagnosis accuracy are improved, the method is suitable for sewage plant equipment fault diagnosis, and the industrial applicability is enhanced.

The invention relates to the technical field of water conservancy projects, in particular to a dam construction period displacement monitoring and safety early warning system and method based on physical information deep learning, and the method comprises the following steps: S1, data preprocessing: carrying out the self-adaptive wavelet denoising and auto-encoder reconstruction of original displacement data, and completing the standardization processing; s2, hybrid model displacement prediction: after training a PINN-LSTM / GRU model through historical monitoring data, predicting future displacement, and using NeuralProphet to correct a prediction residual error; and S3, multi-factor dynamic safety evaluation and early warning: adjusting threshold parameters and weight coefficients in combination with a construction stage, dynamically evaluating a safety risk level, and issuing the safety risk level through a graded early warning signal. According to the invention, high-precision displacement prediction and multi-factor dynamic safety early warning are realized through fusion of physical information and deep learning, and the monitoring reliability and early warning response capability of the dam construction period are effectively improved.