107 results about "Wavelet packet transformation" patented technology

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 discloses a cross-working-condition fault diagnosis method based on knowledge embedding and multi-scale attention, belongs to the technical field of cross-working-condition fault diagnosis, designs a domain knowledge embedding signal processing method based on wavelet packet transformation, envelope spectrum analysis and statistical feature analysis, and constructs a knowledge feature matrix and a statistical feature matrix. Fault information which is not easily influenced by working condition changes is highlighted, and the dependence of a deep learning model on a target domain sample is effectively reduced; a multi-scale attention mechanism is designed to extract depth fault features in an input matrix, and compared with an original multi-head attention mechanism, multi-scale is introduced, so that the parameter quantity of a model is reduced, and the feature extraction capability is more flexible; by embedding domain knowledge highlighting domain invariant fault information into a signal processing end, the method can show excellent variable working condition fault diagnosis performance when a target domain sample is completely lacked.

According to the Mama network solenoid valve fault diagnosis method based on the frequency domain characteristics, the problem of solenoid valve system fault diagnosis can be solved. According to the method, a pneumatic solenoid valve fault data set is constructed by collecting operation signals such as voltage and current, and multi-scale frequency domain features of the signals are extracted by adopting Wavelet Packet Transform (WPT) and Discrete Fourier Transform (DFT). WPT-DFT preprocessing features are embedded into an improved channel-space joint attention mechanism module, and the improved channel-space joint attention mechanism module is combined with a Mamba network with extremely high sequence modeling capability to construct an end-to-end fault diagnosis model. Compared with a traditional convolutional neural network, the method can more effectively capture deep dynamic features in a time sequence and highlight a key frequency region, and experimental results show that the method has higher diagnosis precision and generalization ability and is suitable for intelligent detection of the electromagnetic valve under complex working conditions.

The invention discloses a building safety intelligent monitoring, early warning, prevention and control method, and the method comprises the steps: collecting a physical state parameter and an environment disturbance parameter of a building structure body in real time through a distributed monitoring node group disposed at a key part of the building structure body, and forming an original monitoring data flow; performing space-time alignment and noise reduction processing on the original monitoring data stream by using an adaptive weighted fusion algorithm to generate a standardized structure response data set; extracting a multi-dimensional time-frequency domain feature vector representing the health state of the structure from the standardized structure response data set based on a wavelet packet transformation and principal component analysis combination method; and inputting the multi-dimensional time-frequency domain feature vector into a pre-trained twin neural network, and outputting abnormal region positioning information and an abnormal degree quantitative index. According to the method, the building mechanics mechanism and the artificial intelligence technology are deeply fused, a full-closed-loop intelligent prevention and control system from accurate risk identification to active regulation and control is constructed, and the reliability and timeliness of building safety monitoring in a complex environment are remarkably improved.

The invention discloses a rotor acoustic anomaly detection method, system and device based on an auto-encoder and wavelet packet energy entropy and a medium, and belongs to the technical field of hydroelectric generating sets, and the method comprises the steps: collecting an original acoustic signal, carrying out the noise reduction of the original acoustic signal, and carrying out the multi-channel data fusion to obtain an integrated acoustic signal; inputting the integrated acoustic signal into a wavelet packet for transformation processing to obtain wavelet packet coefficients under different scales; calculating energy values of different frequency band signals based on wavelet packet coefficients to form an energy entropy feature vector; constructing an implicit feature model of the hydroelectric generating set rotor in a normal state according to the energy entropy feature vector; and carrying out feature reconstruction on the energy entropy feature vector, calculating a reconstruction error, carrying out dynamic statistical analysis, and carrying out real-time monitoring and abnormity judgment on the operation state of the hydroelectric generating set rotor. According to the method, the detection precision and the response speed are improved in actual hydroelectric generating set rotor acoustic anomaly detection, and automatic identification and dynamic threshold adaptive adjustment of irregular perturbation are realized.

The invention relates to the technical field of power quality monitoring, in particular to a method and system for constructing a power quality disturbance recognition model based on multidimensional data, and the method comprises the steps: obtaining voltage and current waveform data and environmental parameter data of a key node of a power transmission line, carrying out the hardware defect self-inspection and phase compensation of the voltage and current waveform data, and obtaining a power quality disturbance recognition model; clean transmission electric energy data is obtained; and performing multi-scale noise suppression and time sequence correlation analysis on the clean transmission electric energy data, and constructing a high-fidelity disturbance sequence. According to the method, through the hardware defect self-inspection and phase compensation steps, denoising preprocessing is carried out by utilizing wavelet packet transformation, the frequency response deviation of equipment is identified through Fourier transformation, a frequency domain interpolation method is adopted to reconstruct a frequency-closed defect mark segment, the phase deviation error can be accurately compensated, and the detection accuracy is improved. And self-systematic errors of hardware are eliminated from a data acquisition source, high fidelity of clean transmission electric energy data used for subsequent analysis is ensured, and a foundation is laid for high-precision identification.

The invention provides a data-driven saline screw compressor unit dynamic modeling method, and belongs to the technical field of industrial intelligence and predictive maintenance. According to the method, adaptive deep denoising of original data is realized through a composite signal processing flow of fusing variational mode decomposition, permutation entropy criterion and wavelet packet transformation optimal threshold denoising; meanwhile, a sectional sampling strategy is introduced to enhance the diversity of training data. Then, a wavelet multi-scale energy entropy extraction layer is used for constructing a high-information-density feature vector; furthermore, a prediction model formed by multiple layers of stacked long and short-term memory network units is adopted, and the complex time sequence dependency relationship of the system is deeply captured. According to the method, pure and stable system dynamic representation can be extracted from high-noise industrial data, the dynamic characteristics of the system in the full working condition range are accurately described, and it is ensured that the prediction result is self-consistent physically and reliable in engineering, so that the prediction precision and generalization performance of the model are remarkably improved.

The invention discloses an ultra-high voltage transmission line fault analysis system based on multi-dimensional data, and belongs to the technical field of big data analysis. The method is used for solving the technical problem that the robustness of a system is poor when an existing single-algorithm technical scheme is implemented. Through multi-source data synchronous acquisition and improved EEMD denoising, noise interference of an extra-high voltage line in a complex electromagnetic environment is suppressed; signal characteristics are dynamically tracked through adaptive Kalman filtering, space-time alignment and standardization processing are carried out, dimensional difference and time delay are eliminated, improved wavelet packet transformation, a deep belief network and adaptive morphological filtering are carried out in parallel, multi-dimensional characteristics are dynamically weighted and fused through an attention mechanism, a fault characteristic vector is constructed, and fault diagnosis is carried out. According to the method, efficient extraction and optimization of fault features are achieved, accurate reasoning of fault types and positioning can be achieved through the improved fuzzy Bayesian network, the particle swarm optimization algorithm adaptively updates parameters based on real-time errors, and the adaptive robustness of a complex power grid environment can be effectively improved.

The invention discloses a motor state monitoring method, system and equipment based on multi-sensor data fusion, and relates to the technical field of industrial monitoring, and the method comprises the steps: deploying a flexible strain-temperature composite sensor array at a key position of a motor housing, carrying out the real-time collection to obtain a multi-source fusion signal, carrying out the wavelet packet transformation, and carrying out the real-time collection of the multi-source fusion signal; extracting non-stationary fault fingerprints, calculating gear wear topology invariants, performing incremental parameter exchange with a cloud knowledge base, dynamically generating a health degree confidence ellipse in combination with real-time working conditions, and if the health degree exceeds a threshold value, triggering a brain-like decision-making unit to perform simulation fault-tolerant control. The technical problems that an existing motor state monitoring method is single in sensing data, a feature extraction shallow layer and a decision-making mechanism are solidified, the diagnosis sensitivity of early-stage composite faults is insufficient, and fault-tolerant control lags are solved, and the purposes that multi-physics field games are fused with deep fault fingerprints and cloud incremental evolution are achieved. And the technical effects of prospective identification and real-time fault-tolerant control of potential faults are realized.

The invention discloses an elevator steel belt damage detection and quantitative analysis system based on eddy current and magnetic flux leakage dual-mode detection, and aims to solve the problems of single-mode information loss and serious industrial field strong noise interference in the existing steel belt detection. The system synchronously integrates an eddy current sensor, a magnetic flux leakage sensor and an encoder through a multi-probe array adapter; and multi-dimensional damage physical information in the steel strip is obtained. An improved wavelet packet transform-empirical mode decomposition (WPT-EMD) collaborative noise reduction algorithm is adopted, and in combination with a sub-band energy entropy and a self-attention mechanism, non-stationary mechanical noise is effectively filtered out. A multi-rule feature extraction engine is used for extracting smooth residual errors, derivative mutation and other features, a support vector machine (DE-SVM) model introducing physical priori knowledge weights is constructed, and the small sample recognition problem is solved in combination with a virtual sample generation technology. The method can realize high-precision positioning and quantitative evaluation of steel strip damage under complex working conditions, and has the characteristics of strong anti-interference capability, high identification accuracy and good generalization performance.

The invention discloses a fault diagnosis line selection method based on data driving, and aims to solve the problems that the line selection accuracy is low under complex working conditions such as high-resistance grounding and intermittent arc, and secondary damage of equipment is easily caused by a manual line pulling method in the prior art. The method comprises the following steps: synchronously acquiring zero-sequence current and voltage data of each feeder line; performing multi-scale energy decomposition through wavelet packet transformation; extracting time sequence characteristics of each frequency band by using a parallel gating circulation unit network; constructing a feeder topological graph, and aggregating spatial correlation by adopting a graph convolutional network; and fusing the spatio-temporal characteristics through a self-attention mechanism, and outputting a fault line by a classifier. The system comprises a data synchronous acquisition module, a multi-scale decomposition module, a time domain feature extraction module, a spatial correlation analysis module, a space-time fusion module and a classification module. According to the method, high-precision, high-robustness and non-intrusive fault line selection can be realized without a power grid accurate model and depending on active intervention.

The invention discloses a hidden space adversarial sample generation method and system based on multi-scale feature separation, and the method comprises the steps: employing a neural network quantization training method based on straight-through estimation, and training a hierarchical vector quantization variational auto-encoder; carrying out differentiable Haar wavelet transformation on the input image by adopting a wavelet packet transformation algorithm, decomposing the input image into a low-frequency component and a high-frequency component, and realizing multi-scale feature separation; inputting the high-frequency component into a hierarchical vector quantization variational auto-encoder, and extracting and quantizing global high-frequency features and local high-frequency detail features; in the potential space, a learnable disturbance variable is introduced, a potential vector after disturbance is constructed, and the potential vector is reconstructed into an adversarial sample through a decoder; and based on a preset disturbance target, carrying out iterative optimization on the disturbance vector until a confrontation sample which satisfies an attack success condition and is optimized in visual quality is generated. According to the method, a wavelet domain variational auto-encoder and a hidden space iterative attack algorithm are fused, and an adversarial sample with high fidelity and clear interpretation is generated.

The invention discloses a compressor multi-working-condition feature extraction method based on valve movement acoustic emission, and belongs to the technical field of compressors, and the method comprises the following steps: S1, synchronously collecting compressor valve acoustic emission signals and piston displacement data by using an acoustic emission sensor and a rotating speed encoder, and constructing a time alignment joint signal sequence; s2, calculating a real-time phase according to a piston displacement crankshaft angle, enveloping an energy change rate by combining an acoustic emission signal, and dividing four stages of a compression cycle; s3, extracting the multi-dimensional time-frequency characteristics of the acoustic emission signals in each stage; and S4, inputting the multi-dimensional features into a pre-training auto-encoder, and generating low-dimensional fusion feature vectors to represent operation features. According to the compressor multi-working-condition feature extraction method based on valve movement acoustic emission, acoustic emission and displacement combined modeling is adopted to improve the working condition sensing precision, multi-dimensional features are extracted by means of wavelet packet transformation and spectral analysis fusion to strengthen micro working condition recognition, and the method adapts to multiple typical working conditions and is good in robustness and adaptability.

The invention relates to the technical field of communication transmission, and particularly discloses an anti-electromagnetic interference high-speed transmission data wire harness system and a data wire harness, and the system comprises an electromagnetic environment monitoring module, an interference characteristic analysis module, an interference situation prediction module, a transmission parameter decision module and a transmission execution control module. The method comprises the following steps: acquiring an electromagnetic interference spectrum and transmission quality parameters in real time, and performing feature extraction by using wavelet packet transformation to generate an environment state vector; predicting a future interference situation based on a gradient boosting decision tree model; establishing a dynamic mapping relation between interference intensity and transmission parameters according to a prediction result, and generating a configuration scheme comprising a coding scheme, a power adjustment mode and a modulation mode; parameter switching is executed, transmission quality is verified, and closed-loop control is formed; according to the invention, the technical transformation from passive shielding to active prediction adaptation is realized, the hysteresis problem in coping with complex electromagnetic interference in the prior art is solved, and the reliability of data transmission is improved.

The invention provides a steel structure damage early recognition and positioning method, which comprises the following steps: synchronously collecting and calibrating multi-channel vibration signals, combining sensor space mapping to guarantee data reliability, carrying out band-pass filtering and self-adaptive denoising processing on the signals, and extracting structure local abnormal features through self-adaptive empirical mode decomposition and wavelet packet transformation; a causal inference map is adopted to model a relationship among a damage source, a propagation path and sensor response, a lightweight neural network and a structured attention mechanism are combined, a damage indication factor with physical significance is obtained, and an interpretable damage diagnosis report is output. The method has the advantages of being accurate in damage identification, efficient in feature extraction, high in diagnosis output interpretability and the like, and intelligence and reliability of steel structure health monitoring are improved.

The invention relates to the technical field of brain magnetic signal denoising, and provides a brain magnetic background noise suppression method and device based on multi-scale frequency domain subspace projection filtering. The method comprises the following steps: firstly, converting a multi-channel resting-state noise signal and a multi-channel brain magnetic signal into a time-frequency domain through wavelet packet transformation, and then decomposing data after wavelet packet transformation into different frequency bands so as to separate noise components more clearly; carrying out singular value decomposition on the sub-band coefficient matrix, and adaptively selecting a threshold value by combining an energy accumulation method and a second-order difference method so as to eliminate noise related components; and finally, denoising is performed on each frequency band by using a common subspace projection method, and then the denoised data is reconstructed to obtain the denoised brain magnetic signals, so that the method has a good noise suppression effect, and high-quality clean data can be provided for subsequent brain magnetic signal analysis.

The application relates to the field of alloy defect detection, and specifically discloses a praseodymium-neodymium alloy nondestructive detection method and system based on acoustic feature analysis, which comprehensively captures defect information contained in an original probe signal from two complementary physical perspectives of instantaneous dynamic characteristics and frequency band energy distribution by simultaneously adopting Hilbert-Huang transform and wavelet packet transform. Further, the scheme discards simple feature splicing, and instead utilizes canonical correlation analysis as an information decoupling tool to online decompose two groups of original feature vectors into a shared part describing defect commonality and unique information parts respectively representing the unique resolution capabilities of HHT and wavelet packet. Finally, the three decoupled components are structurally recombined to form a fusion feature vector which can effectively eliminate redundancy, amplify differences and has higher information density, thereby providing a clear structure and highly refined input for a subsequent classification model.

The invention relates to the field of electric reactor fault diagnosis, in particular to an electric reactor mechanical fault diagnosis method based on parameter adaptive optimization, which comprises the following steps: preprocessing vibration signals generated when an electric reactor has a mechanical fault through a wavelet threshold denoising method, extracting vibration signal characteristics through Fourier transform and wavelet packet transform, and calculating the mechanical fault of the electric reactor according to the vibration signal characteristics; according to the method, the high-voltage electric reactor is subjected to fault diagnosis, screening is carried out through a random forest method, fault diagnosis is mainly carried out through a convolutional neural network-Transform hybrid model (CNN-Transform), and accurate diagnosis of the fault of the high-voltage electric reactor is realized by utilizing the local and structured feature extraction capability of the CNN model and the global context and long-distance dependency relationship establishment capability of the Transform model. And an improved dung beetle optimization algorithm (MDBO) is introduced, internal parameters of the hybrid model are dynamically adjusted, parameter adaptive optimization is realized, and the diagnosis accuracy is improved.

The invention relates to the technical field of big data monitoring, in particular to a big data monitoring method and system based on data analysis. The method comprises the following steps: acquiring a high-frequency vibration signal, a rotating speed signal and a temperature signal of to-be-monitored equipment; based on the rotating speed signal and the temperature signal, constructing a dynamic operation load index representing the comprehensive operation load of the equipment; performing wavelet packet transformation on the high-frequency vibration signal to obtain wavelet packet energy of each frequency band; based on the dynamic operation load index, carrying out adaptive normalization processing on the wavelet packet energy so as to decouple equipment working conditions and fault features, and obtaining adaptive normalized wavelet energy; using the adaptive normalized wavelet energy as a feature vector, and using a support vector machine model to monitor and early warn the health state of the equipment; the manual experience dependence and the repeated parameter adjustment cost are reduced, and the data monitoring effect and efficiency are improved.

The application relates to the technical field of FCS-MPC, and discloses a distributed HESS power coordination control method and system based on FCS-MPC, the method comprises the following steps: acquiring power demand P ref that needs to be compensated by an energy storage system ref , using wavelet packet transformation in top-layer control to decompose the P into high power density and high energy density, taking the high power density as a power given value of a super capacitor, taking a high energy density part as a power given value of a storage battery after redistribution, and using multi-step FCS-MPC in bottom-layer control to perform online prediction and optimization control on output follow-up given values of the energy storage units; the system comprises a top-layer power distribution module and a bottom-layer power control module. The application can realize the coordination control of the distributed HESS in an islanded direct-current microgrid, has good control dynamic response performance, stable output voltage, and fast and sensitive power follow-up reaction.

The invention relates to a big data prediction method and device based on machine learning, equipment and a medium, and the method comprises the steps: converting millisecond-level power grid real-time monitoring data into frequency domain features through wavelet packet transformation, and solving the timeliness conflict between high-frequency data and low-frequency commercial data; a tensor fusion technology is utilized to align multi-source heterogeneous information such as power grid topology, enterprise carbon emission and financial data, and unified space-time representation is constructed; structured constraint features are generated in combination with semantic analysis of policies and regulations, and dynamic association between environmental policies and commercial activities is quantified through cross-modal embedding; modeling a risk conduction path based on a dynamic graph neural network, and optimizing a prediction model according to an asymmetric loss mechanism; and finally, outputting a supply chain risk control strategy, and realizing closed-loop response from power grid fluctuation monitoring to supply chain decision. According to the method, the limitation of a traditional model in the aspects of cross-scale data fusion and policy conduction quantification is broken through, and the prediction precision and the response timeliness in an emergency risk scene are remarkably improved.

The present application belongs to the technical field of image recognition, and particularly relates to a liquid crystal screen defect detection method based on image analysis. The method comprises: acquiring a liquid crystal screen surface image frequency domain signal and wavelet packet decomposition, and extracting a low frequency and medium-high frequency coefficient matrix; performing convolution on the low frequency coefficient matrix with an adaptive morphological filter kernel, fitting to generate a uniform illumination reference matrix; after weighted correction and denoising of the medium-high frequency coefficient matrix, inverse wavelet packet transformation is performed on the medium-high frequency coefficient matrix and the uniform illumination reference matrix to generate a reconstructed difference image; the local pixel gradient direction vector and the gray level co-occurrence matrix contrast feature of the reconstructed difference image are input into an isolation forest classifier to output a defect position and category. The present scheme shifts illumination separation to the frequency domain, avoids damage to defect pixels by spatial domain filtering, eliminates non-uniform illumination interference, retains the edge gradient and texture structure of small defects, and solves the problem of missing detection of small dark spots under a gradual illumination background.

The present application belongs to the technical field of compressor abnormal sound detection, and particularly relates to a compressor abnormal sound detection method and system based on acoustic spectrum pattern recognition, which comprises the following steps: extracting a characteristic signal from a sound signal through wavelet packet transform and kurtosis criterion, locating the occurrence time of an impact event contained in the characteristic signal and calculating the time interval of adjacent impact events, calculating a rhythm stability index based on all time intervals in an analysis time window, intercepting a time domain signal window containing each impact event and calculating its energy envelope, calculating the intrinsic damping factor of each energy envelope according to the logarithmic difference between the energy envelope peak value and the energy value after a set time length, calculating the fault authenticity score of each impact event according to the energy value of the impact event in the characteristic signal, the rhythm stability index corresponding to the analysis time window and the intrinsic damping factor, and performing early warning determination, so as to realize high-reliability and accurate early warning of compressor early faults.

The invention relates to the technical field of bridge detection, in particular to a steel-concrete combined bridge damage identification method and system based on sound-vibration fusion, and the method comprises the steps: synchronously collecting vibration and acoustic signals of a bridge, and carrying out the filtering, denoising and standardized preprocessing; inputting the acoustic signal into a noise separation model based on a dual-path recurrent neural network, and separating out pure structural noise; variational mode decomposition and power flow analysis are carried out on the vibration signals, and vibration feature vectors are extracted; wavelet packet transformation and Mel frequency cepstrum coefficient analysis are carried out on the structural noise, and acoustic feature vectors are extracted; after the two types of feature vectors are fused, dimensionality reduction is performed through principal component analysis, a support vector machine model subjected to Bayesian optimization is input, and overall, component and regional multi-stage damage identification and positioning are achieved; according to the method, the advantages of sound and vibration signals are fused, and non-contact, high-precision and intelligent recognition of hidden damage such as interface void, connector degradation and cracks is achieved.

The application discloses an acoustic emission effective signal extraction method, device and equipment and a storage medium, and comprises the following steps: collecting an initial acoustic emission signal, performing frequency band decomposition on the initial acoustic emission signal through wavelet packet transformation, and obtaining a plurality of frequency band waveforms; inputting each frequency band waveform into a preset autoencoder model for reconstruction feature filtering to obtain an initial filtered signal; inputting the initial filtered signal into a preset feature extraction model for time sequence feature filtering to obtain an acoustic emission effective signal; and the preset feature extraction model is used for time sequence feature analysis on the initial filtered signal. Compared with the prior art, the application firstly performs frequency band decomposition on the initial acoustic emission signal through wavelet packet transformation, then inputs different frequency band waveforms into a preset autoencoder model for reconstruction feature filtering, and finally performs time sequence feature analysis on the filtered signal, effectively eliminates background noise, and realizes extraction of the acoustic emission effective signal in a strong background noise environment.

The present application relates to the technical field of intelligent analysis and injury identification of traffic accidents, and discloses a pedestrian-oriented repetitive brain injury risk identification method, which comprises the following steps: constructing an accident database containing human-vehicle and human-ground secondary collision; simulating and restoring the accident scene and extracting head acceleration time history and brain tissue maximum principal strain; performing time-frequency decomposition by wavelet packet transform and calculating wavelet packet energy; linearly weighting and fusing the brain tissue strain of the two collisions by taking the energy proportion as a weighting coefficient to obtain a multi-modal fusion index; constructing a damage probability model based on Weibull distribution, completing parameter maximum likelihood estimation through Bernoulli likelihood function, establishing a damage risk curve, verifying and optimizing the damage risk curve, and finally forming a practical brain injury risk identification curve. The present application comprehensively uses time domain, frequency domain and strain information, quantifies the cumulative damage effect of repetitive collision, effectively improves the comprehensiveness and accuracy of pedestrian brain injury evaluation, and is suitable for vehicle safety design and traffic accident injury identification.

The invention discloses a distribution network relay protection fault detection method and system based on kernel extreme learning, and relates to the technical field of smart power grids, and the method comprises the steps: extracting the three-phase current of a distribution network through wavelet packet transformation features to form a time-frequency energy spectrum; optimizing parameters of the kernel extreme learning machine by using an improved whale algorithm; constructing a new energy power distribution network relay protection fault detection model of the improved whale optimized kernel extreme learning machine, and determining input data of the optimized kernel extreme learning machine and output data of the kernel extreme learning machine; and training and testing the fault detection model to accurately obtain the fault position, type and phase sequence information of the new energy power distribution network. The three-phase current is decomposed into several low-frequency and high-frequency nodes through wavelet packet transformation, information of signals at different frequencies and scales can be provided, the system can better capture the time-frequency characteristics of the current signals, the binary tree is generated to reflect the fault time-frequency characteristics, and the fault diagnosis accuracy is improved. This makes it possible to organize and analyze the time-frequency characteristics of the current signal in a hierarchical manner.

The invention discloses a copper smelting harmonic source positioning and self-adaptive suppression system and method, and belongs to the technical field of copper smelting harmonic positioning. The method comprises the following steps of: acquiring instantaneous voltage signals and instantaneous current signals of equipment in the copper smelting plant by adopting multiple sensors; performing wavelet packet transformation and time-frequency analysis on the voltage signal, extracting a voltage change rate and a phase angle, constructing a spatial disturbance feature vector in a complex form, and calculating a comprehensive harmonic intensity distribution vector by adopting a spatial weighted fusion algorithm; performing support vector machine clustering analysis on the comprehensive harmonic intensity distribution vector, and identifying a position number of a main harmonic source in a power grid topological structure and a corresponding harmonic source contribution degree; constructing a dynamic equivalent circuit model, fitting model parameters in combination with measured data, forming a function model reflecting equipment harmonic injection characteristics, converting a power grid structure into a directed graph, and defining an edge weight as a specific value of a certain harmonic current to a certain harmonic voltage; and harmonic source position information is combined.

The invention belongs to the field of power system protection and control, and particularly relates to a grid-connected and off-grid intelligent switching system of a residual current protection circuit breaker, which comprises a digital twinning construction, safety state evaluation and strategy optimization module, a digital twinning construction module, a safety state evaluation module and a strategy optimization module, constructing a power grid digital twinborn body; on the basis of sliding window detection, wavelet packet transformation and a discrimination model, abnormal current is identified, and a safety state time sequence is generated; and a candidate switching path is generated through a Dijkstra algorithm, the switching opportunity and path are optimized in combination with a particle swarm optimization algorithm, a circuit breaker is driven to execute switching, and data are recorded. According to the invention, grid-connected and off-grid adaptive switching is realized, the power supply safety and continuity are improved, the impact current and interruption time are reduced, and the method is suitable for a distributed energy micro-grid scene.

The invention discloses a hydropower station bus communication quality assessment method, equipment, a medium and a product, and relates to the technical field of industrial bus communication quality assessment, and the method constructs a four-dimensional assessment index system covering a physical layer, a link layer, an application layer and an environment layer for hydropower station multi-protocol and severe environment scenes. Extracting time-frequency domain features of a physical layer through wavelet packet transformation, and combining link layer statistics, application layer semantics and environmental interference features; optimizing a feature fusion weight by adopting improved PCA dimension reduction and a multi-head attention mechanism; and constructing an improved random forest model in which an environment interference weight coefficient is introduced, and carrying out classification evaluation on the feature matrix. According to the method, the problems that in the prior art, evaluation indexes are single, and protocol semantics and environmental interference are not considered are solved, the evaluation accuracy rate reaches 98.5% or above, various communication hidden dangers can be accurately recognized, and the method is suitable for various bus protocols such as Modbus and Profibus DP.