14results about How to "Efficient migration" patented technology

The invention discloses a full-spectrum response BiOBr / Bi2O3-x / WO3-x double-S heterojunction material rich in oxygen vacancy as well as a preparation method and application of the full-spectrum response BiOBr / Bi2O3-x / WO3-x double-S heterojunction material. The heterojunction is prepared by taking bismuth nitrate pentahydrate, potassium bromide and WO3-x microspheres containing oxygen vacancies as main raw materials through a wet impregnation method, and the value range of x in Bi2O3-x and WO3-x is 0 < = xlt; 1 represents that the two have oxygen vacancies with certain concentration. Due to the introduction of the wide-spectrum response semiconductor materials WO3-x and Bi2O3-x, the double-S-type heterojunction shows a remarkable full-spectrum absorption capability. Meanwhile, oxygen vacancies in the material can serve as effective electron capture centers and transmission channels, and separation of photo-induced electron-hole pairs is effectively promoted. Based on the synergistic coupling effect among BiOBr, Bi2O3-x and WO3-x, the heterojunction can degrade ciprofloxacin (CIP) and reduce Cr (VI) (85% and 74% respectively) at the same time. The heterojunction material provided by the invention is simple in preparation process and good in repeatability, and has important application potential in the field of environmental governance.

This invention discloses a 3D seismic fault segmentation method based on 3D adaptive coding and cross-domain collaborative learning, which addresses the problems of scarce annotations, the inability of 2D models to effectively represent 3D structures, and unstable predictions across different work areas. The method first preprocesses the source domain labeled and the target domain unlabeled seismic data. A lightweight 3D adapter is inserted based on the frozen backbone parameters of the 2D visual model, and a dual-path encoder with spatial and depth perspectives is constructed, fusing features from both perspectives to form a joint 3D representation. Then, a lightweight 3D decoder is used to progressively restore resolution and generate a fault probability volume. The source domain uses binary cross-entropy and Dice loss for supervised learning; the target domain uses a heterogeneous bi-sub-network composed of a visual branch and a 3D convolutional branch. Pseudo-labels are adaptively generated through voxel confidence, and unsupervised training is achieved by combining consistency and difference regularization, thus enabling stable and reliable 3D fault segmentation across different work areas even without target domain labels.

The invention belongs to the technical field of solid-state batteries, and discloses an alloy ion / electron mixed conductor and application thereof. The mixed conductor is prepared by taking antimony powder, lithium powder or sodium powder and at least one metal powder of lanthanum, scandium, yttrium, aluminum, gallium and indium as raw materials through ball-milling mixing, tube sealing annealing and quenching treatment. When the mixed conductor is used as an interlayer of a lithium / sodium metal negative electrode and an electrolyte in the solid-state battery, the growth of lithium / sodium dendrites can be effectively inhibited through a specific nucleation mechanism, and the utilization rate of the metal negative electrode is improved; when the conductive additive is used as a conductive additive of a silicon, red phosphorus or graphite / hard carbon negative electrode material, by virtue of high ion / electron conductivity and good potential stability, the interface side reaction is reduced, and the rate capability and the cycling stability of the negative electrode material are remarkably improved.

The present application relates to the technical field of computer vision, and relates to a 3D Gaussian sputtering stylization method, system, device and storage medium; wherein the 3D Gaussian sputtering stylization method comprises the following steps: obtaining a style image and a three-dimensional Gaussian scene to be stylized; performing non-negative matrix decomposition on the style image to extract an initial color basis matrix and an initial coefficient matrix of the style image; constructing a color orthogonal decoding model according to the initial color basis matrix and the initial coefficient matrix, and recalculating color parameters of each Gaussian primitive of the three-dimensional Gaussian scene; performing style transfer optimization on the Gaussian primitive, updating trainable parameters until an iteration termination condition is reached; and determining a stylized three-dimensional Gaussian scene according to updated geometric attribute parameters and color attribute parameters. The present application can solve the problems of color incoordination and dirty color during 3D Gaussian sputtering stylization.

The application discloses a multi-modal low-rank adaptive cognitive function estimation method based on physiological-molecular phenotype, belongs to the technical field of cognitive function estimation, and first collects physiological-molecular phenotype, dynamic physiological signals and cognitive test time information of a subject in a plateau hypoxic environment, and constructs a 28-dimensional multi-modal input feature vector after preprocessing; then, a hybrid expert base model is pre-trained based on a source cognitive task with the largest data volume; finally, the base model parameters are frozen for a new target task, a LoRA low-rank adaptive technology is introduced, only a small number of parameters are trained to complete cross-task adaptation, and the cognitive task reaction time and accuracy are output to realize estimation. The application realizes unified estimation of eight types of core cognitive functions, has high migration efficiency, excellent estimation precision and stability, and can be accurately applied to real-time monitoring of cognitive functions of workers in a plateau hypoxic environment.

The application discloses a bacteriostatic sunscreen self-dehumidifying mask, and relates to the technical field of masks.The bacteriostatic sunscreen self-dehumidifying mask is characterized by a three-layer gradient structure of a hydrophobic layer, a hydrophilic layer and a photothermal layer, synergistic effect of the three layers, efficient directional migration of water under a one-way moisture guiding mechanism, rapid evaporation and self-dehumidification realized in combination with the light-heat conversion characteristics of the photothermal layer, significant reduction of moisture accumulation in the mask, avoidance of facial moisture and bacterial breeding risks during wearing, and improvement of overall comfort and hygiene level.The air vent is located at the lower side to guide the directional circulation of air flow, reduce facial discomfort, and the user can fix the mask through a detachable plastic frame, replace the composite moisture-absorbing gel module conveniently, and ensure long-term and efficient protection.

The invention relates to the technical field of membrane separation, in particular to a chitosan modified MXene electrodialysis membrane as well as a preparation method and application thereof. The invention provides a preparation method of a chitosan modified MXene electrodialysis membrane. The preparation method comprises the following steps: mixing an MXene nanosheet dispersion liquid and a chitosan solution, carrying out a reaction under inert gas protection, carrying out centrifugal washing on the reacted mixed solution, carrying out suction filtration on the obtained dispersion liquid to form a membrane, and carrying out drying treatment. According to the electrodialysis membrane prepared through the method, chitosan molecules are limited between MXene nanosheet layers, a two-dimensional confinement channel structure is formed, the channel regulates and controls the ion migration behavior through the space size and the chemical environment, migration of divalent magnesium ions under the effect of a direct-current electric field is blocked, meanwhile, monovalent lithium ions can migrate preferentially, and the effect of the monovalent lithium ions on the surface of the membrane is improved. Therefore, selective separation of Li < + > / Mg < 2 + > is realized.

This invention discloses a video prediction model compression method and system based on optical flow-guided knowledge distillation, belonging to the field of computer vision technology. The method includes: generating optical flow distillation loss by constraining the optical flow prediction distribution of the student network to be consistent with that of the teacher network; generating channel alignment loss by aligning the channel attention distributions of the teacher and student networks using divergence; generating pixel-level reconstruction loss based on the difference in pixel values ​​between the generated image predicted by the student network and the real image; generating perceptual loss based on the difference in high-level semantic feature space between the generated image predicted by the student network and the real image; and obtaining the trained student network based on the optical flow distillation loss, channel alignment loss, pixel-level reconstruction loss, and perceptual loss. This invention can solve the performance degradation problem caused by neglecting spatiotemporal characteristics in existing compression methods for video prediction tasks.

The invention discloses a gallium nitride device and a preparation method thereof, and relates to the technical field of semiconductor devices. The gallium nitride device comprises a sapphire substrate, an AlN nucleating layer, an N-type AlGaN buffer layer, an intrinsic AlGaN active layer, an electron blocking layer and a P-type AlGaN layer which are sequentially stacked from bottom to top, a gradient doping edge stop layer is arranged on the edge of the P-type AlGaN layer, and meanwhile a bottom electrode in ohmic contact with the N-type AlGaN buffer layer and a top electrode in ohmic contact with the P-type AlGaN layer are formed respectively. According to the invention, by optimizing the structural design of the substrate, the nucleating layer, the buffer layer, the active layer, the electron blocking layer, the edge termination layer and the electrode, and combining with the seagull optimization algorithm to accurately regulate and control the preparation process parameters, the comprehensive improvement of the light extraction efficiency, the carrier recombination efficiency, the breakdown voltage and the long-term working stability is realized; and the photoelectric property and the reliability of the gallium nitride device are obviously enhanced.

The present application relates to the field of environmental monitoring and remote sensing information processing technology, and discloses a method and system for predicting the concentration of nutrient salts in a breeding area coupled with remote sensing data. The method comprises: constructing a physically constrained ecological hydrodynamic model to generate synthetic remote sensing-nutrient salt pairing data; fusing multi-source remote sensing and measured data to form a high-dimensional feature tensor; separating breeding sensitive features and environmental interference features through an attention-guided domain-invariant feature decoupling network; transferring prior knowledge of the source domain large-scale breeding area using a dynamic weight cross-domain knowledge distillation architecture; and finally outputting real-time high-precision nutrient salt concentration distribution through a lightweight regression head and time series smoothing. The system includes corresponding functional units. The present application improves the prediction accuracy and stability in small sample scenarios, suppresses negative transfer, and provides support for precision breeding and eutrophication early warning.

The application relates to the technical field of cloud computing, and particularly discloses a method for cross-cluster migration of a virtual machine, which comprises the following steps: connecting a source cluster where a to-be-migrated virtual machine exists and a target cluster where the to-be-migrated virtual machine is to be migrated to a same backup storage device, generating backup data of the source cluster in the backup storage device, suspending the service of the to-be-migrated virtual machine when the to-be-migrated virtual machine reaches a migration condition, calling a backup recovery interface of the target cluster to recover the backup data to a second back-end storage device mounted by the target cluster, and configuring the to-be-migrated virtual machine in the target cluster, so that the cross-cluster migration of the virtual machine is realized in the form of backup data import and export, the storage structure of the source cluster and the storage structure of the target cluster do not need to be changed, an agent virtual machine for the cross-cluster migration of the virtual machine does not need to be created, and the efficient cross-cluster migration of the virtual machine can be realized. The application also discloses a device and equipment for cross-cluster migration of a virtual machine and a storage medium, which have the above beneficial effects.

The application discloses an illumination self-enhancing photocatalytic material and a preparation method and application thereof, and belongs to the technical field of photocatalytic materials. The material is obtained through cross-linking reaction of a composite photocatalyst, thioctic acid, polyethylene glycol and polyacrylamide, wherein the composite photocatalyst comprises the following raw materials in parts by mass: 1.2-2.4 parts of ammonium molybdate tetrahydrate, 0-5 parts of thiocarbamide and 0.006-0.018 parts of fullerene. The application can realize efficient removal of tetracycline by introducing the composite photocatalyst with high specific surface area and combining with the hydrophilic hydrogel matrix. Meanwhile, the material can enhance the hydrogel strength under illumination, and further improve the mechanical strength and durability of the material.

The invention discloses a pulse reversal-electroosmosis synergistic dechlorination system and method for bridge reinforced concrete, and relates to the technical field of corrosion prevention of reinforced concrete structures. The invention aims to solve the problems of large anode polarization, high steel bar secondary corrosion risk, difficulty in quantification of dechlorination efficiency, large anode rigidity, incapability of fitting a complex structure, no quantitative standard of regeneration efficiency, high anode cost, difficulty in replacement and no derivation basis of service life in the existing constant direct-current electric field dechlorination technology. Through optimization of pulse inversion parameters and electro-osmosis collaborative design, a chloride ion migration amount formula is established, efficient migration of chloride ions is achieved, and the stability of steel bars is guaranteed; the resource circulation value is quantitatively evaluated, and the cost and environment burden are reduced; a dissoluble chitosan frame is designed, an anode service life formula is deduced, automatic disintegration and overall replacement within 24 hours after the anode is used up are achieved, a universal and quantifiable dechlorination system is constructed, and the full demand from efficient dechlorination to low-cost maintenance is met through multi-technology cooperation and formulation evaluation.

The application provides a cross-well area drilling parameter migration method and system based on transfer learning, which first acquires a source well area drilling parameter historical data set, a scene data set, an operation effect data set, and a target well area initial drilling scene data set and operation demand data set, then constructs a cross-well area drilling parameter transfer learning model containing a source domain feature extraction layer, a domain adaptation layer, a knowledge transfer layer and a target domain parameter generation layer, then inputs the source well area data into the source domain feature extraction layer to extract a feature vector, inputs the target well area scene data into the domain adaptation layer to adjust a parameter mapping matrix, then inputs the domain adapted vector and the source well area operation effect data into the knowledge transfer layer to generate a target domain preliminary parameter feature vector, combines the target well area operation demand data to input the target domain parameter generation layer to generate an adapted parameter set, and finally transmits the parameter set to a target well area drilling control system to perform drilling operation, thereby improving the accuracy and efficiency of cross-well area drilling parameter migration.