1718 results about "Pulse sequence" patented technology

The invention relates to the technical field of online calibration, in particular to a multi-parameter fusion intelligent electric energy meter online calibration method and system, and the method comprises the following steps: collecting voltage waveforms, current harmonics and active power data, dividing windows, calculating a covariance matrix, and generating a feature set; a power factor curvature extreme value and a temperature inflection point offset are analyzed to generate an interference identifier, current density distribution and a voltage distortion spectrum are jointly analyzed to extract a harmonic energy ratio to generate a feature vector, and phase compensation is performed on a pulse sequence to generate a calibration instruction set. According to the method, the covariance matrix is constructed by synchronously collecting voltage and current power parameters, the abnormal mark section is generated by combining the temperature change and the covariance difference value, the environment disturbance and the real deviation are effectively distinguished, and the temperature hysteresis effect is identified through the power factor curvature extreme value and the temperature inflection point offset. And combining current density and voltage distortion spectrum analysis to extract a fundamental wave and harmonic wave energy ratio, establishing a composite calibration reference, and dynamically adjusting a pulse duty ratio to realize harmonic wave energy compensation.

The invention discloses a cross-device clock correction method supporting accurate synchronization of meters, and aims to improve the data time consistency among multiple meter devices and meet the requirements of high-precision electric power measurement and event analysis. The method comprises the following steps: acquiring a network time protocol timestamp and a first sampling time sequence in a reference meter, generating a clock reference signal, and broadcasting the clock reference signal to a corrected meter; comparing the corrected meter with a local second sampling time sequence, and calculating a clock skew coarse value vector; further performing wavelet decomposition on the offset coarse value, extracting to form a drift feature vector, fusing the drift feature vector with a historical template, and calculating a unified correction parameter set by using a Bayesian recursion method; and generating a correction pulse sequence carrying a delay residual compensation code based on the parameter set, sending the correction pulse sequence to a corrected meter, calling an interpolation compensation function to construct a piecewise polynomial drift compensation table, and finally realizing dynamic correction and cross-device accurate synchronization of a local real-time clock.

The invention proposes a GIS equipment partial discharge mode recognition method based on deep learning, and relates to the technical field of GIS partial discharge recognition, and the method comprises the steps: collecting a partial discharge signal of GIS equipment through an ultrahigh frequency sensor, and carrying out the signal conditioning; performing digital sampling and processing on the conditioned partial discharge signal to generate a phase-resolved pulse sequence (PRPS) map, and performing intelligent noise reduction processing on the PRPS map, the PRPS map including three-dimensional information of phase, amplitude and frequency; inputting the PRPS atlas into a preset neural network model, and extracting atlas features; and identifying and classifying the partial discharge type based on the extracted map features, and outputting an identification result. According to the method, automatic acquisition, intelligent processing and accurate identification of the partial discharge signals are realized, and compared with a traditional manual identification method, the identification efficiency and accuracy are greatly improved, and reliable technical support is provided for state monitoring and fault diagnosis of GIS equipment.

The invention relates to an intelligent power distribution harmonic monitoring and dynamic compensation system which comprises a monitoring unit, a correction unit and a compensation unit. The monitoring unit continuously collects high-frequency harmonic voltage and current data in a distribution line at a high sampling frequency, extracts transient harmonic components through wavelet packet transformation and empirical mode decomposition, and generates low-dimensional feature vectors based on sparse representation. And the correction unit decodes the low-dimensional feature vector, recovers harmonic time-frequency features, calculates a phase drift rate, predicts a harmonic propagation path and an accumulation node by combining real-time power distribution network topology construction and adopting a nonlinear dynamic prediction model, and generates a correction instruction when abnormality is detected. And the compensation unit adopts pulse sequence density modulation to dynamically adjust a compensation current phase according to the correction instruction, and meanwhile, an inductive coupling device is utilized to transfer harmonic energy to a low-risk node, so that harmonic voltage distortion of a target node is quickly recovered to a stable level in a fundamental wave period after early warning.

The invention relates to the technical field of equipment fault prediction, in particular to a cup labeling equipment fault prediction method and system based on machine learning. According to the method, equipment operation state parameters are converted into a multi-mode pulse sequence with a timestamp synchronization characteristic; loading the purified pulse flow to a quantum bit array for entanglement state evolution, and extracting a three-mode entanglement association tensor; carrying out dimensionality reduction projection on the three-mode correlation tensor to an equipment degradation manifold space, and determining quantum tunneling probability density distribution; constructing a time-varying Hamiltonian of an equipment degradation state based on quantum tunneling probability density distribution, and generating a degradation track cluster according to the time-varying Hamiltonian; and performing time sequence convolution processing on the degradation track cluster, performing probability amplitude amplification on a fault critical point in the track cluster by using an energy level splitting characteristic of a time-varying Hamiltonian, and generating a space-time probability cloud picture. The fault evolution law can be visually presented, the accuracy and timeliness of early fault early warning are improved, and a reliable basis is provided for predictive maintenance.

The invention provides an unmanned cluster brain-like navigation map fusion construction and cooperative positioning method. The method comprises the following steps: acquiring multi-modal perception data, respectively acquiring vision, sound wave and pose information through a binocular camera, a radar and an inertial sensor, simulating a human brain nerve coding mechanism, and generating a pulse sequence and a feature vector in combination with spatio-temporal information; constructing a multi-modal coding unit into a hypergraph node, and based on a dynamic hyperedge connection topological relation, aggregating spatial-temporal characteristics through a heterogeneous hypergraph convolutional network to generate a high-order brain-like semantic map of a single agent; sharing a local brain-like map by multiple agents through distributed communication, detecting geometric and semantic conflicts of an overlapped region, performing space-time alignment based on an anchor point reference, eliminating feature contradictions by utilizing probability distribution matching, and generating a global consistent high-confidence brain-like map; and outputting the optimal position estimation. Through bionic neural coding, heterogeneous hypergraph modeling and multi-agent collaborative optimization, establishment and positioning of a high-order brain-like map in a dynamic unknown environment are realized.

The invention relates to the technical field of trajectory tracking, in particular to an unmanned aerial vehicle trajectory tracking target classification method and system based on a multi-frequency-band radar, and the method comprises the following steps: synchronously collecting original radar echo data of a millimeter wave frequency band, a Ku frequency band and an L frequency band, and generating a multi-frequency-band data set comprising a timestamp, an azimuth angle and a pulse sequence; extracting multi-dimensional features of the target from the synchronized multi-band data set, and generating a multi-band feature vector; constructing a dynamic trajectory association matrix, and determining the space-time continuity of a target trajectory through a trajectory association algorithm based on an attention mechanism; constructing a lightweight classification model; performing online classification on the real-time multi-band feature vectors by using a lightweight classification model, and outputting a target type and confidence; according to the method, capture of the motion state of the target is enhanced, rapid response can be made to tiny motion and change of the target in a complex scene, and therefore more comprehensive and accurate time sequence context information is provided for subsequent classification tasks.

The invention provides an old cell environment sensor data processing method and system based on ultraviolet light communication. The method comprises the following steps: acquiring an environment feature set of an external facade of an old community building through a distributed environment sensor array, and constructing an ultraviolet channel reflection skeleton in combination with an obstacle coordinate; fusing the aging factor attenuation model and the physical space topological structure through a dynamic space-time grid division engine, and generating a multi-dimensional feature segment set carrying space-time attenuation weights; analyzing channel interference in real time based on the interference intensity probability matrix of the ultraviolet channel reflection skeleton, and generating a multi-band transmission configuration result; and calibrating multipath transmission time delay through a phase compensation factor to construct an ultraviolet light pulse sequence carrying a path feature code, and when the difference degree of adjacent units exceeds a critical condition, triggering a closed loop feedback mechanism and updating a link weight. According to the technical scheme provided by the embodiment of the invention, the efficiency and precision of old cell environment sensor data processing can be improved.

The invention relates to the technical field of bushing detection, and discloses a bushing insulation state detection method based on multi-parameter real-time monitoring. The method comprises the following steps: acquiring multi-dimensional monitoring data such as a partial discharge pulse sequence, dielectric loss angle tangent value fluctuation data and surface leakage current density distribution in a bushing operation environment in real time; carrying out insulation defect feature extraction on the data set, and generating an insulation defect distribution map containing a creeping discharge track, an internal air gap position mark and a carbonization channel density index; high-risk and potential defect areas are divided by adopting a dynamic threshold segmentation algorithm, boundary coordinates are determined, and an insulation state evaluation vector containing parameters such as a surface discharge intensity index and the like is constructed according to the boundary coordinates; and inputting the vector into a pre-trained insulation aging prediction model, and outputting a bushing residual insulation life estimated value and a defect development priority sequence. According to the method, all-around accurate monitoring and pre-judgment of the insulation state of the sleeve are achieved, and scientific support is provided for operation and maintenance decision making of the sleeve.

The invention discloses a head-mounted brain body workability assessment method based on a space-time dynamic fusion network, belongs to the technical field of artificial intelligence and human body signal acquisition, and solves the problems that an existing method cannot comprehensively reflect the comprehensive state of an individual, and interaction and potential risks among multiple factors are difficult to reveal. The method comprises the steps that physiological parameter signals are collected based on wearable equipment, and an attention and fatigue prediction result is output by an LSTM prediction model; according to the method, the graph structure data packet is analyzed and processed based on the pre-constructed space-time dynamic fusion network model, the working state signal characteristics of the plateau high-risk operating personnel can be accurately evaluated, the space-time graph neural network (STGNN) can capture the complex relation, the high-dimensional topological characteristics are extracted, and the high-risk working personnel can be accurately evaluated. The dynamic event-driven pulse neural network (deSNN) can convert the pulse sequence features into pulse sequence features so as to describe feature values more accurately, iterative optimization enables the pulse sequence features and the pulse sequence features to cooperate, and the accuracy and reliability of time-space dynamic fusion network model evaluation are improved.

The invention discloses a pulse neural network training method based on adaptive threshold integer activation neurons, and the method specifically comprises the following steps: inputting a neuromorphic data stream or a static image, generating a pulse sequence through time coding, and taking the pulse sequence as the input of a model; in the backbone network, performing down-sampling and extracting multi-scale features by using an SNN module based on LIF neurons activated by an adaptive threshold integer; the visual tasks are divided into image recognition, image detection and image segmentation, and are respectively accessed to different processing modules for processing and outputting; according to the invention, on the basis of a Meta-SpikeFormer element module, the spiking neural network based on LIF neurons activated by an adaptive threshold integer value is used, so that the memory overhead during training is reduced, the network performance of direct training is improved, the application range of the spiking neural network is expanded, and the network is better applied to image classification, recognition and segmentation tasks.

The invention relates to the technical field of image processing, in particular to a super-resolution reconstruction system and method based on a pulse neural network, and the system comprises an image pulse encoder, a pulse feature enhancer and a differentiable pulse decoder. An image pulse encoder simulates a receptive field of a retina by using DoG response, performs adaptive pulse distribution on an input low-resolution image in combination with an image gradient, and converts the low-resolution image into a space-time pulse sequence reflecting a high-frequency region and a low-frequency region in the low-resolution image; a pulse feature intensifier performs coarse-grained and fine-grained structure reconstruction of a low-resolution image on the space-time pulse sequence by using a pulse time sequence dependent plasticity mechanism to obtain a granularity feature pulse; and the differentiable pulse decoder converts the granularity characteristic pulse to obtain a corresponding high-resolution image. According to the method, efficient and low-consumption image super-resolution reconstruction is realized through a bionic retina coding mechanism and a pulse time sequence optimization strategy.

The invention relates to the technical field of solar cell processing, in particular to a laser light-transmitting scribing control method and system for a solar thin-film cell. The control method comprises the steps that a large-size substrate is divided into a plurality of galvanometer view fields through a dynamic flight continuous machining mode, a continuous snake-shaped machining path is generated, the laser machining platform is driven to continuously move at the constant speed, and the galvanometer scanning system is synchronously triggered to start lineation in a splicing area in advance. According to the method, a time domain coupling equation of galvanometer scanning frequency-platform movement speed-laser pulse sequence is established, and dynamic compensation of an optical path and self-adaptive adjustment of energy density are achieved in the continuous machining process. Based on a view field expansion algorithm of a nonlinear kinematics model, the phase deviation of a scanning vector and a motion vector is effectively inhibited, and the geometric continuity of a splicing interface in ultra-long stroke machining is guaranteed.

The invention relates to the technical field of electromagnetic valves, in particular to an intelligent control faucet and a faucet body, the faucet carries a control system, and the system comprises an environment interference suppression module, a gesture recognition processing module, an electromagnetic hysteresis compensation module, a multi-parameter coupling optimization module and a residual magnetism elimination control module. According to the method, the infrared compensation proportion is dynamically adjusted by fusing the environment light intensity and the hand position information, the interference suppression capability is improved, the recognition stability under complex illumination is enhanced, the action consistency is extracted in combination with the gesture track waveform and the dynamic response feature, and the recognition accuracy is improved; electromagnetic hysteresis is dynamically corrected based on real-time current and displacement deviation combined with a temperature factor, motion feedback precision is improved, multi-parameter coupling is fused with multi-physical field data to construct a stable control reference, system sensitivity is enhanced, a reverse pulse sequence is designed by capturing a current attenuation waveform, residual magnetism interference is effectively eliminated, and system performance is improved. And multi-dimensional improvement of identification precision, control stability and response efficiency is integrally realized.

A method for separating liquid-like (T2e) from solid-like (T2G) 1H NMR transverse relaxation times in porous media uses novel pulse sequences together with a 1H NMR spectrometer optimized for geological core samples. The method is applied to obtain 1D T2 distributions and 2D T1-T2 maps in organic-rich chalks for quantification of liquid-like signal (micropore fluids, meso-macropore fluids, fluids dissolved in organic matter, and clay-bound water) and solid-like signal (kerogen, bitumen, and clay hydroxyls). The novel pulse sequences comprise a solid-echo, which detects more solid-like signal than an FID. The method is used for fluid typing in micro / meso-macro pores, clay mineral identification, determination of kerogen content, and quantification of solvent-extracted bitumen versus bitumen expelled from kerogen due to swelling from dissolved hydrocarbons. The method is used to quantify the asphaltene, resin, aromatic, and / or saturate content of bitumen in the rock.

The invention discloses a detection method of a high-speed connector and a related device, and the method comprises the steps: carrying out the plugging motion control of the high-speed connector, injecting a pulse sequence carrying a displacement mark into a differential signal pair in the process, and obtaining the initial reflection data containing a plugging displacement-time index; space-time boundary information of a gap coupling section, an initial touch deformation section and a full-insertion steady-state section is analyzed and determined in a segmented mode through reflection characteristics. And generating a mutual excitation pulse pair in the initial touch deformation section, and executing differential sampling processing by using a thermal-electric coupling effect to obtain a thermally induced group delay sequence and a geometrically induced impedance differential sequence. A dynamic zero-reflection anchor point is generated based on double-sequence fusion, and real-time reference plane migration processing is realized to obtain a calibration impedance curve. And repeating the processes of insertion and extraction, generating an elastic difference curve through displacement-time coaxial difference processing of a bidirectional calibration impedance curve, and positioning an irreversible abnormal section of the impedance. According to the method, the plugging transient impedance change can be continuously monitored, and the early failure risk of the high-speed connector can be evaluated in time.

The invention relates to a radar PRI modulation type identification method based on a neural network and a multi-head attention mechanism, and belongs to the field of radar signal processing. First-order difference is carried out on a radar pulse TOA sequence to obtain sample data of a PRI sequence, a median filter is adopted to carry out filtering processing on the sample data to reduce the influence of missing or false pulses, and a model integrating a multi-scale convolution extrusion excitation neural network, a bidirectional gating circulation unit neural network and a multi-head attention mechanism is constructed. The model can extract data local refinement features and capture data long-term time sequence correlation features, a multi-head attention mechanism is used for integrating global information, more weights are applied to important time steps in a gating circulation unit, the model is trained and tested, and the result shows that compared with other neural network methods, the model has the advantages that the algorithm is simple, and the algorithm is easy to implement. According to the method, the recognition rate of six kinds of radar PRI modulation signals is higher under the condition that the ratio of missing pulses to stray pulses is high.

The invention discloses a conformal antenna component manufacturing system and method based on cross-scale space-time frequency shaping femtosecond laser, and belongs to the technical field of micro-nano machining. The system comprises an ultrafast laser excitation module, a laser pulse time / frequency shaping module, a pulse detection module, an online monitoring module and a focusing high-speed scanning galvanometer module. By means of the system, high-selectivity, large-breadth and three-dimensional precision machining of conformal antenna circuit patterns is achieved, and it is ensured that a substrate is not damaged. The femtosecond laser time-frequency shaping technology is combined in the machining process, layered scanning and accurate removal of the heterogeneous multilayer structure are achieved through an adjustable pulse sequence and a high-speed galvanometer system, and the problems that in the prior art, materials are damaged, and the size precision is low are solved. The device further integrates an on-line depth monitoring function, the machining efficiency and the forming quality are remarkably improved, and the device is suitable for high-precision patterning machining of complex curved surface heterogeneous thin film materials.

The invention discloses an online detection method for small and special motor winding defects based on pulsed eddy current and deep learning, and relates to the technical field of nondestructive testing, original pulsed eddy current detection data containing time domain signals and frequency domain features are collected through a multi-channel eddy current sensor array, and a Maxwell equation constraint network is used for denoising and feature enhancement of the data, so that the detection accuracy is improved. Generating a defect feature candidate set in combination with a multi-scale defect generative adversarial network to obtain target eddy current data, applying a chaos frequency modulation pulse sequence to the target eddy current data as an excitation signal, and collecting a response signal by using a phase synchronization frequency locking technology to obtain enhanced eddy current data; the method comprises the following steps: extracting a physical feature of a sensor, extracting a corresponding electromagnetic feature vector containing the physical feature and a deep learning feature, comparing the electromagnetic feature vector with a preset defect feature library, and determining a defect detection result by combining dynamic excitation response difference. And constructing a dynamically updated defect feature library and combining with a fine tuning network.

An example of a system for delivering neurostimulation may include a programming control circuit and a user interface. The programming control circuit may be configured to generate stimulation parameters controlling delivery of the neurostimulation according to a pulse sequence. The pulse sequence may include a series of neurostimulation pulses and be defined by sequence parameters and one or more modulation functions each modulating an adjustable parameter selected from the sequence parameters. The user interface may be configured to set the pulse sequence to a tonic pulse sequence by determining an initial value for each adjustable parameter and set the pulse sequence to a modulated pulse sequence by selecting one or more adjustable parameters, determining a modulation function for each selected adjustable parameter, and applying the determined modulation function to that selected adjustable parameter to modulate the tonic pulse sequence.

The invention relates to a radar working mode open set identification method and device for a double-path auto-encoder, and the method comprises the steps: inputting a to-be-identified sample into a trained DA network, and obtaining a to-be-identified feature vector; wherein the DA network is obtained by training according to a training data set and a mixed loss function; the training data set is obtained according to an original pulse sequence; the mixed loss function is constructed according to a cross entropy loss function, a center loss function and a reconstruction loss function; calculating the distance between the to-be-identified feature vector and the known category feature center to obtain a first distance; wherein the known category feature center is obtained by inputting a training data set into a DA network for training; and obtaining an identification result of the to-be-identified sample according to the first distance and a distance threshold. According to the method, the distinguishability of different radar working modes is improved through the DA network, and the DA network has good generalization and robustness in a complex non-ideal scene through the mixed loss function.

The invention relates to the technical field of electrical fault detection, in particular to a voltage-sensitive device damage diagnosis system combined with pulse test, which comprises the following modules: an original signal acquisition module for inputting pulse sequences to a voltage-sensitive device, respectively recording voltage response signals and current response signals under each pulse sequence, and generating a response signal set; and performing wavelet decomposition on the voltage response signal and the current response signal based on the response signal set. According to the method, the pulse sequences are input, and the voltage response signals and the current response signals under the pulse sequences are recorded, so that data acquisition covers different working states, and feature missing possibly caused by a single test point is reduced. A wavelet decomposition mode is adopted, a voltage response signal and a current response signal are divided into different frequency intervals, separation of high-frequency and low-frequency characteristics is facilitated, local abnormal signals are not covered by global characteristics, and the capability of analyzing short-time fault characteristics is enhanced.

The invention provides an intelligent brain-like decision-making method, system and device with a body, and a storage medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: receiving a task instruction; analyzing the received task instruction based on a pre-trained large language model, outputting a task description semantic vector, and then carrying out multi-modal data acquisition on the environment to obtain multi-modal perception data; task description semantic vectors are supplemented based on the multi-modal perception data, and task description parameters are obtained; converting the task description parameter into a control signal based on a pre-constructed pulse neural network model, wherein the control signal comprises a plurality of pulse frequencies arranged according to the sequence of the action execution time; and executing corresponding actions according to the sequence of the plurality of pulse frequencies. In this way, the multi-modal perception data and the semantic vectors are combined in the large language model to generate the task description parameters to improve the perception precision of the environment, and the task description parameters of the large language model and the pulse sequence of the pulse neural network are combined to improve the decision response efficiency.

The invention discloses a capacity adaptive coordination control system and method for distributed energy storage charging equipment, and relates to the technical field of charging safety control. The method is used for solving the problems of capacity sudden drop, vehicle charging delay and power grid interference caused by lithium precipitation. The method comprises the following steps: firstly, applying micro-current excitation in a charging pulse turn-off period by using a lithium precipitation characteristic monitoring module, extracting a phase deviation angle in a voltage relaxation curve, judging a lithium precipitation risk and generating a risk tag; then, the risk decision-making module calculates a charge migration proportionality coefficient and a floating coefficient in combination with the battery type and infrared thermal imaging information; then, the demand coupling module fuses the departure moment, the electric quantity gap and the power grid frequency fluctuation to generate a urgency degree weight and a pulse compensation parameter; and finally, the power reconstruction module completes charge migration and pulse sequence reconstruction according to the weight and the compensation parameter, triggers relaxation retest, and constructs a whole-flow dynamic coordination mechanism of the energy storage system.

The invention discloses a motion recognition method and system based on a redox photoelectric memristor, a terminal and a storage medium, and the method comprises the steps: selecting classical motions based on a human body motion data set, extracting time sequence data, and coding the time sequence data into an optical pulse sequence; constructing a reservoir array composed of a plurality of photoelectric memristors, and expanding an optical pulse sequence signal into a high-dimensional state vector; constructing a supervised training model, solving a weight matrix, and constructing a memristive cross array; and outputting an action classification result through simulation domain operation based on the output current multi-path light current signals in combination with the memristor cross array. According to the method, the motion features can be directly fed into a rear-end classification network for action recognition without depending on a complex digital feature extraction algorithm, so that the transmission and processing overhead of redundant data is fundamentally eliminated; a high-efficiency, low-delay and high-robustness hardware solution is provided for real-time and anti-noise motion recognition in scenes such as intelligent monitoring and man-machine interaction.

The invention discloses an abnormity monitoring method and system for a digital pressure transmitter, and relates to the technical field of transmitter monitoring, and the method comprises the following steps: obtaining an original pressure response sequence in a high-frequency pressure pulsation environment, and synchronously recording a sampling clock pulse sequence in an acquisition process, establishing double-track mapping based on the original pressure response sequence and the sampling clock pulse sequence, and forming an input condition with a unified time baseline; and performing instantaneous phase difference extraction on the double-track mapping on the unified time base line. Through double-track mapping, phase difference extraction and nonlinear cross spectrum reconstruction, the stroboscopic aliasing risk is quantified and recognition in advance is realized, and the sensitivity and accuracy of abnormal monitoring of the digital pressure transmitter are improved; meanwhile, the multi-scale risk amplification and reversible time-frequency rearrangement technology is combined, covered high-frequency components are separated and recovered, a real pressure pulsation curve is output after dynamic correction, continuous and stable monitoring is guaranteed, and pressure measurement reliability and operation safety are improved.

A method for ablating tissue by applying at least one pulse train of pulsed-field energy. The method includes delivering a pulse train of energy having a predetermined frequency to cardiac tissue, the pulse train including at least 60 pulses, an inter-phase delay between 0 μs and 5 μs, an inter-pulse delay of at least 5 μs, and a pulse width of 5 μs.

The invention discloses a power battery SOC estimation method based on a pulse neural network model. The pulse neural network model sequentially comprises a time alignment module, a pulse coding module, a pulse feature extraction module and a pulse decoding module. The pulse feature extraction module is formed by connecting a plurality of feature extraction units in series, each feature extraction unit comprises a self-attention layer, a pulse neuron layer and a feedforward neural network, and the self-attention layer captures a long and short term dependency relationship in time sequence data by using a self-attention mechanism; the pulse neuron layer converts the output of the self-attention layer into a pulse sequence, and the feedforward neural network carries out regression prediction according to the pulse sequence. The pulse neural network is used for capturing the dynamic real-time response characteristic of the battery and is lower in energy consumption, the self-attention model captures the global dependency relationship of the SOC of the battery, the pulse neural network and the self-attention model are combined to form a hybrid model which has the time response characteristic and can process a long-time sequence and multi-modal information, and the accuracy and robustness of SOC estimation of the battery are improved.

The invention relates to the technical field of laser deposition control, in particular to an unmanned aerial vehicle metal part ultrasonic field auxiliary laser deposition system which comprises the following modules: a data acquisition module for acquiring a molten pool thermal field signal, a sound field signal and a laser scattering signal, and performing sampling, filtering and synchronous processing to obtain molten pool original data; the micro-flow digital twinning module is used for establishing a molten pool micro-flow digital twinning model according to the molten pool original data and simulating molten pool micro-flow behavior to obtain micro-flow prediction data; and the microflow ultrasonic regulation and control module is used for generating an ultrasonic pulse sequence according to the microflow prediction data and applying the ultrasonic pulse sequence to the molten pool to regulate and control the flow field. According to the invention, through the molten pool microflow digital twinning module and the microflow ultrasonic regulation and control module, real-time modeling and regulation and control of the molten pool microflow structure in the complex geometric part deposition process are realized, the problem that local vortexes, shear layer differences and flow field anomalies cannot be identified in real time in the traditional technology is solved, and the molten pool microflow evolution can be actively reconstructed.

The invention belongs to the technical field of communication navigation, and particularly relates to a covert communication method based on TACAN navigation. Inserting an immediate pulse pair into an auxiliary reference pulse of the TACAN signal emission pulse sequence; and performing channel coding on the subsequent pulse pair for transmitting communication data. The uplink transmission communication data frame of the ground control tower sequentially comprises packet header data, first channel estimation data SINBAL1, first effective data DATA1, second channel estimation data SINBAL2, second effective data DATA2 and protection data from front to back; and the lengths of the two bytes are respectively 6 bytes, 2 bytes, 4 bytes, 2 bytes, 4 bytes and 2 bytes.