32results about How to "Prevent degradation" patented technology

The application discloses a combined power generation method and system based on multi-temperature-zone waste heat recovery. The combined power generation method comprises the following steps: acquiring the cell power of a fuel cell and the power generation power of a hydrogen generator, inputting hydrogen into a multi-stage power generation device to drive the fuel cell to perform a power generation operation, obtaining output first electric energy, first waste heat gas, first heat energy and unreacted hydrogen; inputting the unreacted hydrogen into the hydrogen generator to drive the hydrogen generator to perform a power generation operation according to the hydrogen and the unreacted hydrogen, obtaining second electric energy and second waste heat gas output by the hydrogen generator; acquiring the system temperature of the multi-stage power generation device when performing the power generation operation, and constructing an exhaust passage of an organic Rankine cycle generator to the multi-stage power generation device; driving the organic Rankine cycle generator to perform a power generation operation according to the exhaust passage, the first waste heat gas and the second waste heat gas, and obtaining third electric energy and second heat energy output by the organic Rankine cycle generator, so that the use efficiency of energy is improved.

The application relates to the technical field of electronic reconnaissance and signal processing, and discloses a radiation source individual fine identification method and system based on multi-modal feature fusion, which comprises the following steps: performing preprocessing on obtained radiation source original signals to obtain signal pulse units; extracting time domain sequences of the signal pulse units and performing multi-domain transformation to generate a first time-frequency image, a second time-frequency image and a bispectrum image; inputting the first time-frequency image, the second time-frequency image, the bispectrum image and the time domain sequence into a multi-modal fusion neural network; extracting corresponding image depth features and comprehensive sequence features through three image processing branches and a sequence processing branch; performing cross-modal fusion on the three image depth features and the comprehensive sequence features through a fusion module to output a radiation source individual identification result. The application improves the identification accuracy of radiation source individuals in a complex electromagnetic environment.

The invention relates to a post-treatment method for a mercury-based II-VI group nanocrystal with a protective layer formed by chemical oxidation, and solves the problems of poor stability and photoelectric property degradation of the mercury-based II-VI group nanocrystal caused by surface oxidation, ligand desorption and environmental sensitivity. An oxide protection layer with the thickness of 1-3 nm is formed on the surface of the nanocrystal, the long-term stability and the photoelectric property of the nanocrystal are improved, and the method comprises the specific steps of centrifugal purification of the nanocrystal, preparation of an oxidizing agent solution, controllable oxidation reaction, functional ligand surface modification and purification treatment. The oxidation layer is ensured to be compact and stable by accurately adjusting the concentration of the oxidant, the reaction temperature and the reaction time. According to the method, oxygen and moisture permeation is effectively blocked, surface defects are reduced, the quantum yield and the photoelectric property are improved, and the method is widely applied to the fields of post-treatment of mercury-based II-VI group nanocrystals and photoelectric devices.

The invention discloses a microbial carrier implantable soil microstructure improvement device and method, the device takes microbial carrier precise implantation as a core, integrates multiple functions of targeted layered implantation, carrier-soil microstructure synchronous reconstruction, microcapsule in-situ activation and the like, and can improve the soil microstructure through an intelligent regulation and control system and modular design. And efficient improvement of the soil microstructure in different soil layers and different scenes is realized. The device comprises a power module, a control module and a switchable function module, innovatively adopts a targeted layered injection mechanism, a carrier-soil mixing and shaping assembly and an environmental response type microcapsule spray-seeding activation module, and is matched with a degradable carrier material modification assembly and a new energy driving system. The problems that a traditional device carrier is low in implanting precision, not tightly combined with soil, single in application scene and the like are solved.

A method for producing a dental filling material that includes polyvalent metallic element-including inorganic particles includes a step of treating the inorganic particles with a solution including a chelating agent.

This invention discloses a method for jointly regulating water resource utilization of desert riverbank vegetation based on ecological gate groups. The method includes: obtaining vegetation change patterns using a fitting method based on the kNDVI vegetation index; dividing the vegetation in the desert watershed into three vegetation distribution zones based on the vegetation change patterns and vegetation coverage; acquiring all ecological gates in the desert watershed and dividing them into multiple ecological gate groups; determining the weights of the three vegetation distribution zones for ecological protection objectives using the analytic hierarchy process (AHP), and calculating the comprehensive score of the regulation zone based on the proportion of ecological functional zones in the regulation zone; solving the joint model of the ecological gate groups using a single-objective genetic algorithm to obtain the ecological water demand and ecological water supply of each control zone when the total water shortage of all ecological gate groups is minimized; dividing the ecological gate groups into five levels based on the comprehensive score, and then activating the ecological gate groups to irrigate the irrigation area according to the ecological gate group level, ecological water demand, and the importance of the irrigation area.

This invention discloses an image segmentation method and medium for densely populated live cell regions. The method includes acquiring multiple first data pairs and preprocessing each first data pair. Each first data pair includes a single-modal master image of a live cell region and a paired real auxiliary modality image. A feature extraction network is trained using the preprocessed first data pairs, and the feature extraction network outputs a predicted auxiliary modality image corresponding to the single-modal master image. The preprocessed first data pairs and the predicted auxiliary modality image are input into a pre-trained segmentation model, and the dynamic weights of the predicted auxiliary modality image are adjusted to optimize a second loss function to obtain optimal weight parameters. The single-modal master image of the live cell image to be segmented is input into the trained feature extraction network, and combined with the optimal weight parameters, it is input into the pre-trained segmentation model to obtain the segmentation result. This method is particularly effective for accurate segmentation in densely populated regions.

The application belongs to the field of traffic flow prediction, and provides a city traffic flow prediction method, system and product under complex road network conditions, and the technical scheme is as follows: time mixing features are extracted based on city road network traffic flow time series data, intersection information between different traffic features is captured based on the time mixing features, mixed feature representation is obtained by fusing the time mixing features and the intersection information between different traffic features; global dependence in a long sequence is captured by selective state space modeling based on the mixed feature representation, and dynamic time series features are output; after the dynamic time series features are enhanced, a trailing time dimension is projected to obtain prediction values of each feature variable of the traffic time series data. In city traffic flow prediction, data can be dynamically processed according to real-time traffic conditions, multivariate information can be effectively utilized, and efficient long sequence modeling can be performed, and the prediction accuracy of time series data is significantly improved.

This application discloses a few-shot target detection and recognition method based on self-supervised subspace separation, belonging to the field of remote sensing image processing technology. The proposed innovative framework completely isolates the training of base classes and novel classes at three levels. First, a low-rank subspace adapter is proposed for structural separation, achieving network optimization for novel classes without compromising the pre-training performance of base classes. This uses fewer parameters to modulate network components, alleviating the overfitting problem caused by fully fine-tuning under few-shot conditions. Second, an orthogonal subspace extractor is proposed for feature decoupling, adaptively learning the corresponding subspace for each class and extracting decoupled representations, enhancing the separability between classes. Third, a balanced classifier is designed for loss balancing, preventing the final prediction result from being overly biased towards the background or base class. The framework proposed in this invention has significant advantages in few-shot target detection and recognition, showing significant improvements in both base and novel classes.

The application relates to the technical field of artificial intelligence aided design, and discloses a man-machine cooperation visual design method and system based on multi-dimensional literacy grading, which collects literacy features from four dimensions of technical operation, content decoding, creative application and innovation integration and vectorizes the features, constructs an adaptive grading portrait by using Gaussian mixture clustering, generates a cooperation strategy based on multi-target Pareto optimization, drives man-machine cooperation closed-loop iterative design, and realizes continuous self-optimization through multi-dimensional quality evaluation and literacy increment feedback.

The invention discloses a core muscle group unbalance evaluation and active rehabilitation control method and system in a weightless environment, and the system part comprises an underwater bioelectricity sensing module, an embedded evaluation and control module, and an active rehabilitation execution module. An underwater core muscle group unbalance evaluation and control strategy based on surface myoelectricity feature extraction and principal component analysis dimensionality reduction is adopted, and the strategy comprises the steps that firstly, bioelectricity signals of a trunk core muscle group are collected through a waterproof sensor; extracting time domain, frequency domain and nonlinear features, and calculating a lateral asymmetry index and a front-back balance index; reducing the dimension of the high-dimensional index into two core components, namely an unbalance amplitude and an unbalance mode, by using a PCA model; judging the current physiological state according to the scores of the two components; and finally, controlling the underwater work clothes or the exoskeleton to execute differentiated rehabilitation actions (amplitude remodeling or mode induction) on the basis. And in the rehabilitation intervention process, continuously and circularly monitoring until the muscle group control indexes return to the baseline range.

This invention discloses an inverse perovskite solar cell and its fabrication method that utilizes a tetrathiofulvalene derivative to regulate the self-assembled monolayer (SAM) assembly structure. The invention uses bis(carbazolium phosphate) SAM-4PABCz as the hole-selective interface layer, employs an ethanol / NMP mixed solvent system, and introduces a tetrathiofulvalene (TTF) derivative as a dopant. By leveraging the steric hindrance effect and non-covalent intermolecular interactions of the TTF derivative, the out-of-plane π-π interactions between bis(carbazolium phosphate) molecules are weakened, inducing the molecules to shift from a high-tilt-angle "standing" stacking of SAM molecules to a low-tilt-angle near-flat arrangement. This regulation strategy effectively suppresses island-like aggregation of SAMs, reduces the interface defect state density, promotes uniform and ordered molecular stacking, and forms a uniform and dense hole-selective contact interface.

The application provides a visible light-infrared pedestrian re-identification method and system, and belongs to the technical field of computer vision. Details and contours are extracted by strengthening details and contours according to the characteristics of visible light images or infrared images, and then denoising by mutual calibration of modes, bidirectional optimization of features, to obtain optimized features with cross-modal consistency and discriminability, to generate double-mode enhanced features, to solve the problem of insufficient feature extraction granularity and modal adaptability, and to avoid visible light texture loss or infrared contour fragmentation. Meanwhile, the application also avoids modal specific information degradation through mutual calibration of modes, strengthens cross-modal semantic alignment, and improves the representation capacity of the model and the robustness to complex scenes.

Provided is a Low-E glass plate protective sheet having excellent long-term reliability. A Low-E glass plate protective sheet is provided. The protective sheet has a PSA layer. The PSA layer has a chloride ion amount of 105 µg or less per gram of PSA, determined by hot water extraction.

This invention relates to a SiC VDMOSFET based on a non-floating buried layer and its fabrication method, comprising: an N+ substrate; an N-drift region located on the N+ substrate; a P-type non-floating buried layer extending from the surface of the N-drift region to its interior; an epitaxial N-region located on the N-drift region; a trench region extending from the surface of the epitaxial N-region to its interior, with its bottom separated from the P-type non-floating buried layer by a predetermined distance; a P+ source region extending from the bottom of the trench region until it overlaps with the P-type non-floating buried layer and is formed on the sidewall of the trench region; a Pwell region extending from the surface of the epitaxial N-region to its interior, contacting the P+ source region on the sidewall of the trench region; and an N+ source region extending from the surface of the Pwell region to its interior and contacting the P+ source region. This invention significantly improves the long-term gate oxide reliability and overall high-frequency switching performance of the device without sacrificing device symmetry or effective conduction area.

The application discloses a text question answering method and device, equipment and a computer readable storage medium, comprising: receiving a text question to be answered of any length; finding an answer to the text question to be answered by using a target question answering model obtained by iterative training of a loss function with gradient truncation, to obtain a target answer; wherein the loss function with gradient truncation comprises a stop gradient operator for blocking the gradient of the model output item corresponding to the intermediate state text in the back propagation; each sample pair in the training sample set used for training the target question answering model comprises a core clue sample, an intermediate state sample and a full sample; and an output operation is performed on the target answer. The application improves the direct alignment effect and can balance the question answering performance of short texts and long texts.

This disclosure relates to the field of face recognition technology, and provides a training method and apparatus for a face recognition model that avoids the long tail problem of data. The method includes: constructing a face recognition model; obtaining a training dataset, and executing the following loop to train the face recognition model in multiple rounds: sampling the current round of training from the training dataset using a dynamic sampler to obtain a sample set used for the current round of training; inputting the sample set into a feature extraction network to obtain a feature vector set corresponding to the sample set; inputting the feature vector set into a normalization network to normalize the feature vectors in the feature vector set; calculating a classification loss using a classification network and a contrastive loss using a contrastive network based on the feature vector set processed by the normalization network; updating the model parameters of the face recognition model based on the classification loss and contrastive loss; incrementing the training round number corresponding to the current round of training by one; and ending the loop when the training round number equals a preset round number.

A back contact cell and a preparation method thereof, the back contact cell comprising a silicon wafer, a dielectric layer, a P-type passivation layer, a protruding passivation layer and a P-type electrode, the dielectric layer being arranged on a back light surface of the silicon wafer; the P-type passivation layer and the protruding passivation layer are both arranged on a side of the dielectric layer away from the silicon wafer; wherein the thickness of the protruding passivation layer is less than the thickness of the P-type passivation layer, the protruding passivation layer is formed of a passivation material, and the protruding passivation layer and the P-type passivation layer are in electrical contact; the P-type electrode is electrically connected to the P-type passivation layer, thus, the present application can improve the open circuit voltage and the short circuit current, while ensuring the long-term reliability of the ohmic contact between the electrode and the emitter.

The invention discloses a combined power generation method and system based on multi-temperature-zone waste heat recovery, and the method comprises the steps: obtaining the power of a fuel cell and the power generation power of a hydrogen power generator, inputting hydrogen into a multi-stage power generation device, and then driving the fuel cell to execute the power generation operation, output first electric energy, first waste heat gas, first heat energy and unreacted hydrogen are obtained; inputting the unreacted hydrogen into a hydrogen generator, and driving the hydrogen generator to execute power generation operation according to the hydrogen and the unreacted hydrogen to obtain second electric energy and second waste heat gas output by the hydrogen generator; the system temperature of the multi-stage power generation device during power generation operation is obtained, so that an exhaust channel from the organic Rankine cycle power generator to the multi-stage power generation device is constructed; and the organic Rankine cycle generator is driven to execute power generation operation according to the exhaust channel, the first waste heat gas and the second waste heat gas, third electric energy and second heat energy output by the organic Rankine cycle generator are obtained, and the use efficiency of energy is improved.

This invention discloses an electrolyte for use under high cutoff voltage conditions and its preparation method, relating to the field of battery electrolyte materials technology. This electrolyte, through extraction using a reference electrolyte solution, can specifically extract electrochemically active molecular components from anthocyanins, allowing them to preferentially adsorb onto the negatively charged positive electrode surface under high cutoff voltage conditions via electrostatic interactions. This not only interrupts the chain reaction but also directly anchors lattice oxygen, inhibiting lattice oxygen release at its source. It achieves a synergistic effect of a dual protection mechanism of "dynamic removal and static anchoring." Furthermore, high-speed centrifugation avoids increased ion transport impedance. This electrolyte significantly inhibits chain reactions under high cutoff voltage conditions, effectively preventing the formation of excessively thick or high-impedance interface layers, ensuring the battery's capacity and cycle capacity retention under high cutoff voltage conditions, making it suitable for large-scale application in lithium batteries under high cutoff voltage conditions.

The present application relates to the technical field of multi-modal data prediction, and discloses a fashion preference prediction method and device, which comprises obtaining multi-modal data including images, texts and user behavior time sequence, performing feature extraction to obtain image features, text features and time sequence features; constructing spatial features based on city embedding tables and corresponding regional culture label codes; aligning the dimensions of the spatial features and the time sequence features, fusing the aligned spatial features and the time sequence features by using a gated attention mechanism to obtain spatio-temporal fusion features; splicing the spatio-temporal fusion features with the image features and the text features to obtain multi-modal fusion features; decoupling the multi-modal fusion features to obtain material decoupling features, style decoupling features and scene decoupling features; and after feature splicing of the various decoupling features, performing Transform coding and MLP classification to obtain fashion preference prediction probability.

Provided is a fuel cell electrode catalyst in which catalytic metal particles made of Pt or a Pt alloy are supported on a carbon carrier, wherein the number of the catalytic metal particles per unit surface area of the carbon carrier is 4.3 / 100 nm2 to 16.0 / 100 nm2.

The invention discloses a connector and a processing method. The connector comprises a shell; a plurality of contact elements are arranged on the shell in an array penetrating manner; the shell comprises an extending part and a connecting part, and the connecting part forms a bulge towards the outer side of the extending part; the end faces, close to one side of the extending part, of the extending part and the protrusion of the shell are first non-plating areas, and the end face of the connecting part and the side face of the protrusion are first plating areas. The first plating layer area is locally selected for plating nickel, palladium and gold on the connector shell to cover the insertion contact welding area of the connector and the T / R assembly interface, so that the conductivity, weldability, wear resistance and corrosion resistance of the contact area of the connector and the T / R assembly interface are improved, the dielectric pressure resistance is improved, and the manufacturing cost is effectively controlled. The housing and the contact element are integrally connected, so that heat generated during working of the T / R assembly can be quickly led out, performance degradation of an interconnection interface of the connector caused by high temperature is effectively avoided, and stable operation under a high-power-density working condition is ensured.

The application discloses a kind of open-pit mine road network generation methods based on two-way graph convolution network, belong to open-pit mine road network generation field.Region sample division is carried out to open-pit mine track area, and then the region sample is grid divided;Multi-track feature representation of each grid is generated using track data in the grid, and region grid track feature map is obtained;A road centerline prediction model comprising a residual encoder network, a two-way graph convolution network and a decoder network is constructed, and a road centerline probability map is generated using the region grid track feature map;The road centerline probability map is converted into a road centerline map, and all predicted road centerline maps are spliced to generate an initial road network for the open-pit mine, and the topology of the initial road network is refined by connecting broken road edges, and finally the open-pit mine road network is obtained.The application can extract rich track features, enhance the description of roads, improve the perception of spatial information and channel information, and improve the integrity and continuity of the road network topology.

The invention discloses a barium titanate ceramic capacitor material and a preparation method thereof, the composition of the barium titanate ceramic capacitor material is represented as (1-a-b-c-d-e) BaTiO3-aNb2O5-bGd2O3-cB2O3-dSiO2-eMn3O4, a = 0.008-0.012, b = 0.002-0.005, c = 0.001-0.004, d = 0.001-0.004, and e = 0.002-0.004, the preparation method comprises the following steps: (1) mixing Nb2O5, Gd2O3, B2O3, SiO2 and Mn3O4 according to a ratio, carrying out mixing, ball milling and sieving, and then carrying out first sintering treatment; (2) adding BaTiO3 into the material prepared in the step (1), mixing and ball-milling, and drying; and (3) carrying out compression molding on the material obtained in the step (2), and then carrying out secondary sintering treatment. The barium titanate ceramic capacitor material disclosed by the invention realizes low loss and high temperature stability while keeping relatively high dielectric.

The invention discloses a human immature testis in-vitro culture contamination model and a construction method thereof, and belongs to the technical field of biomedicine. According to the method, the human testis tissue is directly used and is closer to the physiological and toxicological response of the human body, the testis beside the benign testis tumor are taken, and a sample is easier to obtain; the culture at 34 DEG C better conforms to the physiological environment of the testis, and tissue degeneration at 37 DEG C is avoided; providing two models, namely 10 <-5 > M (4-day model, accelerated toxicity characterization for mechanism screening) and 10 <-7 > M (10-day model, prolonged contamination time and real reaction of time accumulation effect of toxicity); multi-level evaluation systems of morphology, molecular biology, transcriptomics and the like are integrated; the method can be used for testicular toxicity screening and mechanism research of other environmental toxicants.

The invention relates to an aqueous pharmaceutical composition comprising: a) bilastine or a pharmaceutically acceptable salt or solvate thereof, b) mometasone, or a pharmaceutically acceptable derivative thereof, c) a suspending agent, and d) 2-hydroxypropyl-3-cyclodextrin; the pH of the aqueous pharmaceutical composition being between 3.5 and 5.5, and the content of 2-hydroxypropyl-β-cyclodextrin being less than 8.5% by weight. The invention also relates to said compositions intended for use in the treatment and / or prevention of a disorder or disease that may improve through antagonism of the histamine h1 receptor and / or a disease responsive to corticosteroids by nasal administration. The invention further relates to a method for preparing the aforementioned aqueous pharmaceutical composition.

The application discloses a high-small-bark-cocaine-content high-yield-type new variety of Coptis chinensis breeding method, relates to the Coptis chinensis breeding technical field, and aims at solving the low yield of Coptis chinensis. The application sets a small-bark-cocaine-content hard threshold of 5.5%, combines with actual planting related indexes such as agronomy and stress resistance, and comprehensively screens germplasm, so that the screening standard is more definite, and high-quality germplasm can be accurately screened. From field cultivation to multi-dimensional identification to three rounds of reselection, the logic is closed, the operation is standard, and the high-yield, high-quality and stable Coptis chinensis excellent single strain can be effectively directionally bred.

The application discloses a high-precision three-dimensional vision system based on multi-camera time-sharing acquisition to improve acquisition frequency, and relates to the technical field of three-dimensional vision, comprising: a time-sharing synchronous control module, which is used for generating and outputting a group of trigger signals with precise time sequence offset; an image acquisition unit array, which is composed of at least three spatially distributed image acquisition units, each unit responds to the trigger signal, and sequentially opens its image sensor for image acquisition within a unified system acquisition cycle, and the acquisition windows of each unit do not overlap in time; a time-sharing acquisition cooperative scheduling module, which is used for dynamically solving and assigning the mutually exclusive acquisition window parameters of each image acquisition unit within the acquisition cycle according to scene motion speed, illumination change and system state. The application solves the problem that the existing three-dimensional vision system is difficult to consider high frequency and high precision use scene.