56results about How to "Good practical engineering application value" patented technology

The invention relates to a geological disaster prediction method and system based on an optimized PU-XGBoost model, and belongs to the technical field of geological disaster monitoring and risk assessment, and the method comprises the steps: obtaining multi-source data including InSAR deformation monitoring data and historical geological disaster point data, taking the historical geological disaster points and InSAR significant deformation points in the InSAR deformation monitoring data as a positive sample set P, and carrying out the prediction of the InSAR significant deformation points in the historical geological disaster points; a PU-Learning strategy of a Spy sample is introduced to construct a reliable negative sample system, an XGBoost hyper-parameter is optimized and searched in combination with introduction of a BTO optimization algorithm, an optimal hyper-parameter combination is obtained, and an optimal model is obtained through training of a positive sample set P and a negative sample set RN; and performing geological disaster susceptibility prediction and hierarchical mapping on the research area based on the optimal model, thereby realizing high-precision and high-reliability evaluation of the geological disaster susceptibility of the research area.

The invention relates to a high-precision steel surface defect detection method suitable for a complex industrial environment, and aims to solve the problems of feature coupling and background interference caused by data scarcity, long tail distribution and multi-defect coexistence. According to the method, a conditional generative adversarial network (CGAN) and a convolutional neural network (CNN) are combined, an SE-Net channel attention mechanism is integrated, and the method is used for automatic detection and classification of multiple defects on the steel surface. By introducing a physical constraint CGAN data generation method, a multi-defect coexistence composite image conforming to industrial reality can be generated, so that the diversity and accuracy of a training data set are effectively enhanced. The model optimizes the expression of defect features through multi-level feature extraction and an attention mechanism, and enhances the distinguishing ability and recognition precision of multiple defects. Experiments show that the method is excellent in performance in a multi-defect identification task, the accuracy rate of 98.89% and the F1 score of 99.72% are achieved, the method is remarkably superior to a traditional detection method and other deep learning models, and the method has high robustness and practical value in a complex industrial environment.

This invention discloses a multi-frequency, multi-target range-velocity deambiguation method and system based on phased array fire control radar, mainly addressing the low accuracy of existing methods in handling multi-target ambiguity. The implementation scheme is as follows: acquiring measurement points from the phased array fire control radar and generating a full-space candidate point set; discretizing the range-velocity plane into a grid, counting the number of candidate points and non-repeating frequency groups within each grid block; dividing the grid into blocks based on block-level neighborhood statistics to generate the candidate point set; calculating the unique frequency support number of the candidate point set and performing clustering to obtain candidate point clusters; statistically estimating the candidate point clusters, outputting target range and velocity estimates, forming the deambigued target set for the current frame. This invention can reduce the probability of false clustering structures formed by erroneous unfolding points, clutter points, and false alarm points, improving the accuracy and real-time processing efficiency of multi-target range-velocity joint deambiguation in complex scenarios, and can be used for multi-target signal processing and target tracking in phased array fire control radar.

The application discloses a kind of millimeter wave broadband filtering antennas with wide-angle beam scanning function and array, from top to bottom include broadband radiator, metal ground layer and microstrip line feed layer in sequence, the broadband radiator includes first rectangular patch, second rectangular patch and driving patch;By controlling the size of first metal patch, second metal patch and driving patch, the resonant frequency of broadband radiator is regulated;By adjusting the distance between first rectangular patch, second rectangular patch and driving patch, and the spacing of adjacent first metal patch and adjacent second metal patch, the coupling amount of broadband radiator is regulated, and the bandwidth of broadband radiator is further regulated.The application has simple and compact structure, wide impedance bandwidth, and can cover four millimeter wave frequency bands.

The invention discloses a muon imaging detection system and method based on scintillation fiber and SiPM reading. Comprising at least one detection unit, the detection unit comprises a scintillation optical fiber array and SiPM array support plates, the scintillation optical fiber array is formed by arranging a plurality of scintillation optical fibers in parallel, the two ends of the scintillation optical fiber array are each provided with one SiPM array support plate, SiPM arrays are integrated on the SiPM array support plates, the number of the SiPM units is the same as that of the scintillation optical fibers, and the SiPM units and the scintillation optical fibers are in one-to-one correspondence; the SiPM unit is optically coupled with one end of a corresponding scintillation optical fiber signal, receives a scintillation optical signal transmitted in the corresponding scintillation optical fiber and generates an electric signal; the signal processing unit is used for preprocessing the electric signal output by the SiPM to obtain a processed signal; and the positioning reconstruction unit is used for reconstructing the space motion trail of the muon according to the processed signal. The system can realize three-dimensional high-precision positioning and track reconstruction of muons, and has the advantages of high integration level, modular design and the like.

A very low frequency circuit frequency response parameter calibration system is suitable for high-precision frequency response calibration of a signal processing circuit working in a frequency range of 300 Hz to 100 kHz. The system comprises a signal acquisition module and an FPGA. The signal acquisition module comprises three synchronous acquisition channels, two output channels of a to-be-calibrated circuit, and a GPIO outputting a calibration signal of the FPGA. The FPGA is configured to generate a calibration signal after receiving a configuration instruction of an upper computer system. The calibration signal is divided into three paths: one path enters a very low frequency signal acquisition module as a reference signal for calibration comparison; the other two paths are injected into two input channels of the to-be-calibrated circuit to drive the to-be-calibrated circuit to generate corresponding outputs; the three sampling channels of the very low frequency signal acquisition module are synchronously controlled to start data acquisition and complete isochronous sampling tasks, ensuring the synchronization of the three paths of data in the time domain; after sampling is completed, the three-channel data is preprocessed, and then the preprocessed data is uploaded to the upper computer system for visual processing.

This invention discloses a vehicle trust assessment method for cooperative decision-making at unsignalized intersections, belonging to the field of intelligent transportation and autonomous driving cooperative control technology. The method first acquires multi-source state and interaction information of the target vehicle and conflicting vehicles, constructing reference position estimates and physical interaction characteristics. Then, based on multi-source observation consistency analysis, it calculates the information reliability of the target vehicle and, combined with the spatiotemporal interaction relationship between vehicles, calculates the physical interaction trust of the conflicting vehicle pairs. Further, it determines the comprehensive trust based on the information reliability and physical interaction trust, and generates a safe time interval between vehicles based on the comprehensive trust. This interval is then used as a conflict constraint parameter input into the intersection cooperative scheduling model to achieve adaptive optimization control of vehicle passage order and entry time. This method can adaptively adjust the safe time interval between vehicles under conditions of uncertainty in vehicle state information, differences in vehicle execution consistency, and complex interaction relationships, improving the safety, stability, and traffic efficiency of cooperative decision-making at unsignalized intersections.

The application provides a variable driving parameter silicon carbide MOSFET active driving method, circuit and device. The active driving method comprises: in the on process, a power supply is controlled to apply a first weak driving level in a first time period and a first strong driving level in other time periods, the starting time of the first time period is the time when the silicon carbide MOSFET enters the Miller platform, and the ending time is the time when the drain current reaches the current overshoot peak value; in the judgment process, the power supply is controlled to apply a second weak driving level in a third time period and a second strong driving level in other time periods, the starting time of the third time period is the time when the drain-source voltage is equal to the DC bus voltage, and the ending time is the time when the drain-source voltage reaches the voltage overshoot peak value. The method has good versatility and high engineering application value, reduces the additional switching loss caused by weak driving, maintains high switching speed while ensuring the overshoot suppression effect, and thus realizes a better compromise between overshoot, loss and switching time.

A non-intrusive multi-load load monitoring method and system based on shared time-series coding and load-specific refinement is disclosed, relating to the fields of smart grid, non-intrusive load monitoring, and energy data analysis. This method addresses the lack of existing non-intrusive multi-load load monitoring methods that can simultaneously address shared modeling, load difference representation, training stability, and engineering deployment feasibility. The process involves generating a fixed-length input window; after the input window is compressed to form latent space temporal features, multiple shared time-series coding modules extract the shared time-series representations for the corresponding multiple loads; these shared time-series representations are then fed into multiple independent load-specific refinement heads, which generate power predictions and state predictions for the target loads; residual gating is applied to the power predictions of the target loads using the state predictions, yielding the final decomposition results for the multiple target loads.

The invention discloses a device and a method for testing anisotropic mechanical properties of strong heterogeneity rocks, the device comprises a rotatable rock sample fixing device, a workbench and a sliding track, and a scratching cutter control and data monitoring system, the workbench and the sliding track device main body are fixed on a table top and are convenient to support and move; the slidable scratching system moving track supports the slidable scratching system to move left and right; the rotatable rock sample fixing device is located on the left side above the table top and can fix the position of a rock sample and change the rock scratch angle. The scratching tool control and data monitoring system is located on the right side of the table top and connected with the sliding rail, the slidable scratching system can scratch rocks, and the data monitoring system can monitor and collect tangential force data borne by a tool bit and displacement data of the tool bit in real time; the device and the method for testing the anisotropy of the shale based on the scratches can be researched so as to judge the anisotropy mechanical property of the strong heterogeneity rock.

The application discloses a high-parallelism storage-computation integrated polar code decoding method based on coset decoding, and belongs to the technical field of channel coding and decoding. The method comprises the following steps: constructing an error pattern group and a check constraint subgroup; performing coset division on the error pattern group to obtain a set of mutually disjoint cosets; screening a limited number of minimum cosets based on the minimum Hamming weight of the cosets; and performing parallel decoding processing on the paths corresponding to the minimum cosets in the storage-computation integrated architecture. k The number of paths in the traditional multi-bit decoding is reduced from 2 to a preset number L through coset division, and the parallel execution of the likelihood information updating and the path metric calculation is realized in combination with the calculation-type storage array, so that the decoding performance is ensured, the calculation complexity and the hardware resource consumption are significantly reduced, and the decoding throughput and the energy efficiency are improved.

This invention discloses a multidimensional indexed modulation chaotic communication method and apparatus based on neural network-assisted detection, belonging to the field of wireless communication and chaotic spread spectrum communication technology. The method includes: firstly, segmenting the input bitstream to obtain modulation bits, domain index bits, codeword index bits, and reference index bits; then, selecting the corresponding transform domain codebook based on the domain index and the corresponding codeword basis vector based on the codeword index, while simultaneously generating an orthogonal chaotic sequence pool and selecting the corresponding reference chaotic sequence; next, performing MPSK modulation on the modulation bits, and combining the codeword basis vector and the reference chaotic sequence to generate a multidimensional indexed modulation chaotic signal, which is then transmitted through a wireless channel after multi-carrier modulation; the receiving end demodulates, filters, and samples the received signal, and inputs it into a multi-task convolutional neural network detector to achieve joint detection of modulation symbols, domain index, codeword index, and reference index, thereby recovering the transmitted bitstream.

The application discloses a construction method of a drag-reducing isolation structure applied to a precast pile, and belongs to the technical field of pile foundation construction, which comprises a precast pile, a drag-reducing isolation material, an isolation protective layer, a cement-soil pile, an energy circulation channel and an energy medium pump; the drag-reducing isolation material is uniformly distributed on the surface of the precast pile, the isolation protective layer is wrapped outside the precast pile and the drag-reducing isolation material, the cement-soil pile is outside the isolation protective layer and is bonded between the isolation protective layer; the energy circulation channel is arranged inside the precast pile, and the energy circulation channel and the energy medium pump are connected through a pipeline. The construction steps are as follows: precast pile manufacturing; drag-reducing isolation material manufacturing; isolation protective layer manufacturing; drag-reducing isolation material smearing; isolation protective layer wrapping; cement-soil pile construction; implanted precast pile processing; energy circulation reduces and eliminates negative friction drag; and continuous maintenance. The application can reduce the influence of negative friction drag on the precast pile through multiple active drag reduction, and has good engineering application value and popularization value.

The invention relates to alumina toughened zirconia ceramic powder as well as a preparation method and application thereof. The invention develops a modification preparation method which integrates precise proportioning, multi-stage PH regulation and control, composite coating agent modification and controllable aging process, and definitely adopts a solid-to-liquid ratio of (50-200): (100-1500) (the solid-to-liquid mass ratio is (0.5-2): (1-15)) to ensure the stability of the slurry, so that uniform dispersion of powder, strong interface bonding and collaborative optimization of mechanical properties and chemical stability are realized. Compared with the prior art, the invention has the advantages of good powder dispersibility, strong phase interface bonding, high slurry stability, excellent ceramic mechanical properties and chemical stability, and the like.

The application relates to the technical field of wind power equipment decommissioning and recycling, in particular to a decommissioned wind power blade on-site modular efficient disassembling method. First, the decommissioned wind power blade is subjected to structure scanning to obtain internal main beam area and shell structure distribution information; then, an unfoldable temporary bearing framework is introduced into the blade interior to form a temporary bearing system in the blade interior; under the condition that the temporary bearing system continuously supports, the blade shell is subjected to longitudinal unloading cutting and circumferential modular cutting; then, the bonding interface between the main beam and the shell is subjected to softening treatment, and the expansion separation mode is used to realize directional decoupling of the main beam structure and the shell structure; finally, each module formed by disassembling is turned over, transported and classified and recycled. The application can maintain the overall force stability of the blade during the disassembling process, reduce the structural collapse risk, improve the modular disassembling efficiency and the composite material classification and recycling integrity, and has good engineering application value.

The invention particularly relates to a multi-jacket liquid nitrogen ice plug freezing method using a non-uniform gap structure, and belongs to the technical field of nuclear power plant pipeline freezing, and the method comprises the following steps: step 1, determining the total length L1 of jackets and the number N of the jackets of a pipeline freezing section; 2, calculating the value range of the total gap length L2 of the jacket of the frozen section of the pipeline; and step 3, designing and distributing non-uniform gaps. By designing the non-uniform jacket gaps, energy regulation and control are carried out in the ice plug forming process, so that the initial freezing position is accurately controlled, the complete freezing time is shortened, the pressure bearing capacity of the ice plug is improved, and finally energy conservation and efficiency improvement are achieved.

This invention provides a method for recognizing the operating status of solar thermal collectors based on a structure-aware segmentation network, belonging to the field of solar thermal equipment detection technology based on computer vision. First, it acquires equipment images and video data under different operating conditions through multi-source visual acquisition methods, and constructs a multidimensional dataset containing structural semantic labels, status labels, and temporal labels. Then, it builds a structure-aware semantic segmentation model based on an improved DeepLab network, and achieves high-precision segmentation of key structural regions of the thermal collector by introducing structural prior coding and structural constraint attention mechanisms. On this basis, it combines a temporal consistency discrimination module to model structural changes in continuous video frames, achieving dynamic recognition of states such as occlusion, offset, damage, and abnormal installation angles. Finally, it fuses the segmentation results and status discrimination results to output the equipment operating status and anomaly type, realizing automated detection and analysis.

The application provides a near-field electromagnetic source positioning method based on array signal processing, relates to the technical field of electromagnetic signal detection and positioning, and comprises the following steps: constructing an array receiving signal model based on the spherical wave front characteristics of electromagnetic waves, wherein a steering vector is used in the construction; collecting near-field radiation source signals by using a sensor array; and performing amplitude calibration processing on the near-field radiation source signals to form an array data set. The near-field electromagnetic source positioning method based on array signal processing realizes high-precision three-dimensional positioning of a near-field electromagnetic source by combining amplitude calibration, characteristic decomposition and joint optimization processing with the array receiving signal model based on the spherical wave front characteristics of electromagnetic waves. The near-field electromagnetic source positioning method based on array signal processing introduces a multi-dimensional optimization strategy in the signal subspace extraction and parameter estimation process, improves the estimation accuracy of the distance, the pitch angle and the azimuth angle, and improves the positioning stability and robustness in a complex electromagnetic environment.

This invention discloses a laser-arc flexible heat source multi-degree-of-freedom real-time adjustable intelligent welding torch and a real-time dynamic control method. The arc welding torch mounting assembly of this invention is compatible with both TIG and MAG welding torches. The laser head mounting assembly and the arc welding torch mounting assembly are mounted on the main support of the welding torch, and a multi-degree-of-freedom electrified control mechanism is used to precisely adjust the distance and angle between the laser heat source and the arc heat source in real time and dynamically. The laser vision sensing system is mounted in front of the intelligent welding torch to detect the weld trajectory and bevel assembly accuracy in real time, and relies on the welding quality closed-loop control system to issue welding process control commands to the electrified control mechanism of the intelligent welding torch and the welding equipment. This invention realizes the electrified flexible control of the spatial position and angle of the laser-arc flexible heat source, which can flexibly construct laser-TIG or laser-MAG flexible welding heat sources, significantly broadening the welding process window under complex working conditions and improving the intelligence level of the welding process.

This invention relates to the field of foundation pit construction safety monitoring technology, and provides a method and system for foundation pit monitoring analysis and geotechnical parameter inversion. The method includes: building a finite element model of the foundation pit to generate a batch of simulation monitoring curve datasets and binding labels; preprocessing the simulation monitoring curve data and field data respectively, and performing logarithmic transformation on the permeability coefficient labels of soft clay and confined aquifers; constructing and training a two-level integrated model composed of a classification neural network and multiple sets of regression subnetworks; inputting field data into the trained model for two-level inference parameter inversion; performing an inverse exponential transformation on the permeability coefficient to restore the true value and performing risk classification and early warning. This invention can replace measured samples with simulation data, improve inversion accuracy through a two-level integrated architecture, and realize intelligent monitoring and parameter inversion of complex strata.

A kind of urban rainwater pipe network and river network hydrodynamic accelerated calculation method based on data-driven surrogate model, comprising: based on bidirectional long short-term memory (BiLSTM), graph attention network (GAT) and residual connection, data-driven surrogate model Deep-Drain is constructed, for simulating the simplified process of urban rainwater pipe network and river network water flow;Mechanism model is constructed to simulate one-dimensional hydrodynamic process under measured rainfall and different return period design rainfall, and process data set of rainwater pipe network and river network flow process is constructed, while training set, validation set and test set are divided;Optimal hyperparameter combination of Deep-Drain is found by using Bayesian optimization;The efficiency improvement effect of Deep-Drain is quantitatively analyzed by typhoon storm event;The present application realizes the effective improvement of the calculation efficiency of urban rainwater pipe network and river network flow simulation process.

This invention relates to the field of hot extrusion manufacturing technology, specifically to a high-strength weldable 6xxx series aluminum alloy profile, its manufacturing method, and its application. The aluminum alloy profile is composed of the following weight percentages: Si: 0.80%~1.2%, Fe≤0.20%, Cu: 0.40%~0.80%, Mn: 0.30%~0.70%, Mg: 0.70%~1.1%, Cr≤0.25%, Zn≤0.10%, Ti≤0.10%, Zr: 0.05%-0.20%, Pb<0.01%, with the balance being Al and unavoidable impurities. The profile meets the following performance requirements: tensile properties, R... m ≥420MPa, R p0.2 ≥330MPa, A 50mm ≥8.0%; Fatigue performance: at r=0.1, N=1×10 7 Under the same conditions, the fatigue strength σ of the smooth sample is ≥145MPa, realizing the lightweight manufacturing of high-speed trains.

The application discloses a kind of micro-nano robots magnetic gradient force calculation method and system under three synchronous rotating magnetic field.The method comprises the following steps: obtaining the position parameters of three synchronous rotating permanent magnets, the magnetic moment size of permanent magnet, common rotating phase, the position parameters of the space point to be calculated and the magnetic moment size of micro-nano robot;Based on magnetic dipole model, the magnetic field generated by three permanent magnets at the space point to be calculated is calculated and superimposed to obtain the total magnetic field;According to the total magnetic field, the total magnetic field direction unit vector is constructed, and the magnetic moment direction of micro-nano robot is kept consistent with the total magnetic field direction;Field matrix, magnetic gradient matrix and gradient force compact matrix are constructed, and the magnetic gradient force vector and its module length suffered by micro-nano robot at the space point to be calculated are calculated.The application can quickly complete the magnetic gradient force calculation under the action of three synchronous rotating magnetic field, and is suitable for trajectory planning, closed-loop control and safety evaluation of magnetic driving micro-nano robot.

The invention discloses a multi-modal emotion recognition method based on a dynamic variance perception mechanism (DVAM) and truncation penalty, and belongs to the technical field of artificial intelligence and multi-modal emotion calculation. According to the method, through five core steps of multi-modal feature extraction and alignment, DVAM, graph modeling and cross-modal interaction, an index truncation similarity penalty mechanism (ECSP) and feature fusion and sentiment classification, the problems of insufficient dynamics, noise redundancy, obvious modal dominant effect and insufficient cross-modal difference modeling in the prior art are solved. DVAM realizes time sequence variance dynamic perception and weighted modulation of multi-modal features, ECSP realizes adaptive constraint of cross-modal feature similarity, and graph modeling and multi-granularity feature fusion are combined to significantly improve robustness, modal balance and emotion recognition accuracy of multi-modal fusion. And on IEMOCAP and MELD data sets, weighted average F1 values respectively reach 75.42% and 68.35% and are averagely improved by about 2.7% compared with a mainstream fusion model, so that the method has a good engineering application value.

The application relates to the field of intelligent analysis, and discloses a LIBS multi-component quantitative analysis method based on spectrum mechanism constraint and feature decoupling. Target spectrum data is acquired and pretreated. Spectrum mechanism perception features are extracted according to the pretreated target spectrum data. Component-specific feature decoupling is performed according to the spectrum mechanism perception features. Adaptive error regulation is performed. Finally, multi-component quantitative analysis results are output. The output multi-component quantitative analysis results are based on concentration range constraint and multi-component consistency constraint. The application realizes spectrum feature decoupling, avoids the problem of insufficient attention to low-content components, and improves the accuracy and stability of analysis.

The application discloses a kind of based on flight attitude multi pesticide tank unmanned aerial vehicle centroid balance liquid supply control method, obtain unmanned aerial vehicle pitch angle data in flight process, each pesticide tank geometric parameter and liquid state information, according to the change of flight attitude, pesticide tank in liquid free surface form is classified, calculates each pesticide tank liquid centroid coordinate and unmanned aerial vehicle whole machine centroid coordinate. 0-1 mixed linear model is established, under the premise of meeting continuous liquid supply time length constraint, pesticide tank mutual exclusion liquid supply constraint, joint liquid supply amount constraint, liquid supply rate constraint and the liquid amount constraint required by liquid spray system, with the minimum unmanned aerial vehicle whole machine centroid offset of each pesticide tank liquid supply operation as target, dynamically determine the liquid supply state and liquid supply rate of each pesticide tank, realize the collaborative liquid supply control of multi pesticide tank.The application with initiative control and prediction model will instantaneous centroid always be constrained in ideal small range, effectively reduce the centroid deviation of multi pesticide tank unmanned aerial vehicle, improve the flight stability of plant protection unmanned aerial vehicle.

This invention relates to an aluminum-based composite material, a robot arm, and a method for preparing the robot arm, belonging to the technical field of metal matrix composites and high-end equipment structural components. The aluminum-based composite material comprises an aluminum alloy matrix and silicon carbide particle reinforcement distributed within the aluminum alloy matrix. The chemical composition of the aluminum alloy matrix, by mass percentage, includes: Cu 3.0-6.0%, Mg 0.5-2.0%, Ag 0.1-1.0%, Zr 0.05-0.3%, Ti 0.01-0.2%, with the balance being Al and unavoidable impurities. The microstructure of the aluminum alloy matrix contains L12-Al3(Zr,Ti) nanophases and lath-like Al2Cu phases. The robot arm prepared using this aluminum-based composite material exhibits an axial tensile strength exceeding 320 MPa and an axial thermal conductivity exceeding 220 W / (m·K) at a high temperature of 300℃, achieving an integrated function of "load bearing and heat conduction." This invention effectively solves the problems of decreased accuracy and shortened lifespan of existing robot arms due to material softening and insufficient heat conduction at high temperatures, and has the advantages of lightweight, high reliability and low cost.

The invention belongs to the technical field of temperature field simulation and data-driven reconstruction, and relates to a thermodynamic information guided cyclic diagram network cutting tool temperature field reconstruction method. Firstly, heat flux is determined based on heat transfer characteristics in the cutting process of the tool. Parameter design is conducted on heat flux through a Latin hypercube sampling method, a cutter heat transfer finite element model is established, temperature field data under different parameter conditions are obtained, and sample data are provided for model training. Then, a graph structure model is constructed according to the node space topological relation, a gating circulation unit is introduced to describe the time evolution characteristic of the temperature field, and a temperature field reconstruction model is established by combining physical consistency information such as global maximum value keeping and local extreme value constraint; model parameters are trained through an optimizer, so that the model can learn the spatial correlation and time dependency relationship of the temperature field. And for an unknown space node or an unknown thermal load state, performing temperature prediction by using the trained model to realize three-dimensional temperature field information reconstruction.