28results about How to "Strong application value" patented technology

The invention discloses a scale and structure consistency image generation method fusing three-dimensional geometric prior, and belongs to the technical field of computer vision and image generation. The method comprises the following steps: acquiring a scene image and depth information corresponding to the scene image, reconstructing a scene point cloud containing pixel space coordinates and color information, acquiring a colored point cloud of an object to be inserted, placing the colored point cloud at a specified position of the scene through rigid body transformation, and unifying a coordinate system with the scene point cloud; projecting the transformed object point cloud to an image plane through a camera model, determining a visible point according to the depth, and constructing an insertion region mask to remove pixels of a region corresponding to a scene; then constructing a multi-channel geometric structure priori graph containing color, depth and other information, and splicing the multi-channel geometric structure priori graph with the cleared scene image to form a diffusion type generation model input; and finally, generating an image under geometric prior guidance, ensuring that the inserted object is consistent with the original scene in the aspects of position, scale, perspective and the like, solving the problem of insufficient geometric consistency of the existing method, and realizing image content insertion with an accurate structure and real vision.

An ultra-wideband end-fire antenna with a common aperture embedded in a metal carrier includes a metal cavity with a coaxial cable compartment on one side. A radiator module is connected to the bottom of the metal cavity via a radiator fixing slot. The radiator module includes a common aperture antenna substrate. A low-frequency radiator is printed on the side of the common aperture antenna substrate closer to the coaxial cable compartment, and a high-frequency radiator is printed on the side farther from the coaxial cable compartment. This invention utilizes a dual-band antenna in a common aperture design to enhance the channel capacity and anti-fading capability of the communication system. At the same time, by achieving the antenna's carrier conformal characteristics, high isolation, and miniaturization, it enriches the antenna's application scenarios and improves the overall performance of the communication system.

This invention discloses a method for controllable preparation of nanoparticles using a multi-channel microreactor. The method utilizes a microchannel reactor to segment droplets and prepare quantitatively synthesized nanoparticles with controllable particle size. The invention employs capillary action to draw the reaction solution into the microchannel, segmenting one or more droplets (each droplet into 10-100 portions). By changing the width of the microchannel, controllable preparation of nanoparticles within a small range (±3 nm) can be achieved; by changing the length of the microchannel, controllable preparation of nanoparticles within a large range (±10-30 nm) can be achieved. Furthermore, the reactor structure of this invention is simple, enabling highly precise and quantitative synthesis of various nanoparticles. Compared to large-scale reactors, the miniaturization of the synthesis reaction in this invention effectively reduces energy waste and safety hazards, making it highly valuable for applications in safe chemical engineering, green chemical engineering, high-efficiency chemical engineering, microchemical engineering, and precise reaction monitoring.

The invention belongs to the technical field of wireless communication, and particularly relates to a multi-user symbiotic wireless communication channel estimation and signal detection method. According to the designed transmission frame structure, the main system sending symbol is recovered by estimating the combined channel containing the reflection device sending symbol, and then the multiplicative relationship between the main system sending symbol and the reflection device sending symbol is decoupled, so that the error rate performance of main system transmission and reflection device transmission is improved. The scheme provided by the invention is simple to implement and has very high application value.

The application discloses a verifiable privacy protection convolutional neural network system, which uses a given trained convolutional neural network on private data, characterized in that a private verifiable scheme or a public verifiable scheme is adopted. The application provides a verifiable privacy protection convolutional neural network system, which realizes data privacy and verifiability at the same time. The system provides two schemes to meet the needs of different application scenarios, both of which realize data privacy and verifiability. The difference lies in that one is a private verifiable scheme with faster verification speed but needing a private key in the verification process, and the other is a public verifiable scheme with slower verification speed but not needing a private key in the verification process. The system disclosed by the application has strong application value.

The invention relates to the field of special films, in particular to a polypropylene-based ultrathin special film and a preparation method thereof. The film is prepared from the following components in parts by weight: 80 to 90 parts of isotactic polypropylene resin, 10 to 20 parts of syndiotactic polypropylene resin, 5 to 10 parts of a composite modifier, 5 to 10 parts of a compatilizer, 3 to 8 parts of amino modified nano silicon dioxide, 0.1 to 0.5 part of an antioxidant, 0.3 to 0.7 part of a lubricant, 0.1 to 0.3 part of an anti-hydrolysis agent, 0.1 to 0.3 part of a nucleating agent and 0.05 to 0.3 part of a light stabilizer. The special film prepared by the preparation method disclosed by the invention can still keep excellent mechanical bearing capacity under the ultrathin specification with the thickness of less than or equal to 20 microns, and the mechanical property, the flame retardant property, the weather resistance, the processability and the interfacial compatibility are synergistically improved.

The invention discloses a cascade dam group system service risk matrix type evaluation method, which comprises the steps of collecting data information of a drainage basin where a cascade dam group system is located to construct a deep learning model, and performing dam break flood routing calculation; calculating a system failure risk rate of each unit dam, quantifying a risk transfer coefficient between adjacent unit dams, constructing a state transition probability matrix and a state transition loss matrix of the cascade dam group system, calculating a state transition risk matrix of the cascade dam group system, and obtaining a total risk rate of the system and three types of total risks of life, economy and social environment; the corresponding total loss is evaluated; and finally carrying out cascade dam group system risk rating. The method has higher calculation efficiency and precision, and can provide richer and more comprehensive calculation results; according to the system risk quantitative grading standard and method, theoretical preciseness is considered, meanwhile, higher engineering applicability and practical operability are achieved, and higher practical application value is shown.

The application provides a method for realizing well logging lithology identification based on a convolutional recurrent deep learning network of a self-attention mechanism, mainly comprising: (1) data set construction, that is, using drilling core lithology description results and well logging data of a coring well section to group samples; (2) model construction, that is, determining a sample input mode and a basic structure of the model; (3) model training, that is, removing null values and abnormal values, normalizing sample characteristic values, encoding label values, and dividing the samples into a training set, a verification set and a test set; (4) model evaluation, that is, realizing model construction, parameter optimization, model training and verification based on PyTorch. The application has reasonable conception, and realizes well logging lithology identification by constructing a core-well logging mapping relationship.

The application discloses an image coding, decoding and compressing method and system based on vector quantization entropy modeling, which comprises the following steps: obtaining a first feature map of an image to be coded; obtaining a super-prior binary code stream and super-prior information of the first feature map; quantizing the first feature map by using a lattice vector quantizer; projecting the quantized first feature map to an integer coefficient domain of the lattice vector quantizer to obtain a second feature map; modeling the distribution of the second feature map as a Gaussian distribution independent of each dimension, and predicting the mean and variance of the Gaussian distribution by combining the super-prior information and a spatial context model; dividing the integer coefficient domain of the lattice vector quantizer by using a relaxed boundary, and performing probability estimation and arithmetic coding according to the predicted mean and variance of the Gaussian distribution to obtain a feature binary code stream; and combining the feature binary code stream with the super-prior feature binary code stream to obtain a compressed image binary code stream. The application improves the rate-distortion performance of image compression by using a more efficient lattice vector quantization on three-dimensional features.

The application discloses a method and a treatment system for reducing and recycling biological sludge based on catalytic cracking reaction, and comprises the following steps: adding diluent and scale inhibitor into the dewatered biological sludge of a sewage treatment plant for conditioning treatment, so that the form of the biological sludge is changed from solid state into quasi-solid state, and the conditioned biological sludge is obtained; the conditioned biological sludge is transferred into a catalytic cracking reactor through fluidization, and an oxidizing agent is added for catalytic cracking reaction; the biological sludge is partially released into free water through a controllable oxidation process, so that the form of the biological sludge is changed from quasi-solid state into liquid state, and the organic matter in the biological sludge is decomposed into C2-C6-based carboxylic acid, and a carboxylic acid mixture is obtained; the carboxylic acid mixture is treated through heat exchange and cooling, and then is subjected to solid-liquid separation, so that solid residues and filtrate are obtained, and the biological sludge is reduced and recycled at the same time.

The application discloses a process modeling and fault detection method and system for guiding machine learning by control theory, and the method comprises the following steps: acquiring historical operation data of an industrial process, extracting operation dynamics in the historical operation data by using a neural network, and approximating an observer gain part of a kernel representation model to realize complete kernel representation model design; constructing an adjoint system of the kernel representation model by using Hamilton system theory, and guiding and optimizing a learning and training process of the kernel representation model and the adjoint system by using lossless constraints to realize design of a regularized kernel representation model; constructing detection statistics by using residuals of the kernel representation model and the regularized kernel representation model, setting a maximum value of the detection statistics when the industrial process normally operates as a fault detection threshold, and thus realizing fault detection on the industrial process to guarantee safety of the industrial process. In addition, by designing an online fine-tuning strategy, the generalization and applicability of the model are improved, and the method has strong practical value.

The application discloses a kind of defect sample generation method, device, automatic optical detector and storage medium.The defect sample generation method includes: the normal sample output after being detected with AOI detection equipment as research object, and using component positioning method to extract solder joint image, by analyzing the characteristics of defect sample solder joint area, the qualitative characteristics of defect are converted into the quantitative characteristics convenient for modeling analysis, and the complete scheme of defect sample generation is provided using graphic modeling and image processing technology.The generation of a variety of defect samples is realized, and it is convenient to batch generate defect samples, and has strong innovation and application value.

The application discloses a kind of based on multi-modal data fusion's injury condition early warning method and device, method includes: obtaining continuous physiological waveform data and relevant structured data;After waveform data is preprocessed, it is converted to two-dimensional image representation by time series topology enhancement algorithm;Structured data is standardized to extract structured features;Through heterogeneous feature correlation weight mechanism, two kinds of features are semantically aligned and complementary fusion;Based on multi-window risk aggregation algorithm, output injury condition early warning result.The application solves the problems of difficulty in waveform and structured data fusion, insufficient use of time series features, and poor early warning stability in the prior art by enhanced image representation of waveform data, deep multi-modal fusion, and multi-scale risk aggregation, improving the accuracy, generalization ability, and engineering practicality of injury condition early warning. It can be widely applied in clinical first aid, intensive care and trauma treatment scenarios.

The application discloses a moisturizing device for photoresist nozzles and a control method thereof. The device comprises a moisturizing box base, a plurality of nozzles and an ultrasonic spray assembly assembled on the moisturizing box base. The interior of the moisturizing box base is provided with an internal fluid channel for limiting the flow direction of fluid. The internal fluid channel comprises, in sequence, a spray main guide groove, a plurality of spray branch ports and a plurality of nozzle moisturizing ports corresponding to the plurality of spray branch ports. The spray main guide groove is in communication with the ultrasonic spray assembly, and the plurality of nozzles are respectively suspended in the corresponding nozzle moisturizing ports. The application utilizes the ultrasonic assembly to generate windless moisturizing mist. After the mist is slowed down and buffered through the main guide groove and the branch ports, the mist independently enters the nozzle moisturizing ports, respectively, and a high-humidity microenvironment is constructed around each nozzle, which effectively inhibits the volatilization and crystallization of photoresist solvents and solves the problems of nozzle blockage and wafer contamination caused by the volatilization and crystallization of photoresist solvents.

The application discloses a sparse unmixing method based on hypergraph denoising and superpixel weighting, and comprises the following steps: step 1, hypergraph smoothing filtering and denoising of a hyperspectral image; step 2, constructing a spatial weighting matrix based on superpixels; and step 3, weighted sparse unmixing of the hyperspectral image. The method combines hypergraph learning and hypergraph smoothing filtering and denoising to obtain a denoised image, and can effectively preserve the spatial structure of the original image; the denoised image is segmented by using the SLIC superpixel segmentation method, the superpixel structure of the image can be more accurately extracted, and the method has better robustness to Gaussian noises with various signal-to-noise ratios. Therefore, the application has the characteristics of strong practical application value and good reliability, and is suitable for promotion and use.

The application discloses a method and system for identifying farmland runoff pollution in surface water by using microbial fingerprints. Microbial samples of multiple types of water bodies are collected, and microbial composition data are obtained through 16S rDNA sequencing; based on a sensitivity-specificity screening index, combined with specific anaerobic characteristics, a farmland runoff pollution microbial fingerprint is determined, a double-model integrated judgment mechanism is established: an artificial neural network is used to process microbial category data, an XGBoost model is used to analyze microbial relative abundance data, and a control variable method is used to extract the classification decision basis of the double models; and the double models are used to make result judgment based on logical rules. The application innovatively integrates microbial fingerprints and machine learning classification models, can realize accurate and simple identification of farmland runoff pollution in surface water, and improves the identification accuracy and efficiency of farmland runoff pollution.

The invention discloses a multi-spectrogram lightweight acoustic scene classification method and device for auditory auxiliary equipment, equipment and a storage medium. The method comprises the following steps: firstly, acquiring a hearing-aid front-end environment audio signal to be processed, converting the hearing-aid front-end environment audio signal into a Mel spectrum feature map, a constant Q transformation spectrum feature map and a first-order difference spectrum feature map, and then stacking the Mel spectrum feature map, the constant Q transformation spectrum feature map and the first-order difference spectrum feature map to form a multi-channel spectrum feature map containing complementary time-frequency information as model input; then differential shallow layer feature extraction is carried out through an asymmetric multi-stream module, and depth features are extracted on a low-power-consumption computing platform by using a lightweight backbone network with a ghost bottleneck module and coordinate attention as cores; a spectral morphology module, a frequency band attention module and a channel weight balance fusion module are introduced, and multi-dimensional enhancement and dynamic weighting are carried out on the depth features; and finally, outputting an audio scene classification result by the classifier. According to the method, the classification performance of complex acoustic scenes can be improved while the model parameter quantity and the operation power consumption can be remarkably reduced.

The application relates to a comprehensive evaluation method for communication anti-interference performance and belongs to the technical field of communication. The application adopts various algorithms such as data standardization, grey correlation degree calculation and weight calculation, unifies indexes of different anti-interference systems or schemes to one dimension, and calculates the advantages and disadvantages. The application introduces data correlation characteristics into the comprehensive evaluation of the index system, can objectively evaluate the advantages and disadvantages of the comprehensive performance of different systems through a mathematical method, avoids the uncertainty of subjective scoring in the traditional method, and significantly improves the accuracy and reliability of the comprehensive performance evaluation.

The invention relates to the technical field of computer security, in particular to a detection method and device for spreading malicious codes on a picture platform, a storage medium and equipment. When it is detected that a process downloads a picture from a picture platform, data analysis is carried out on the picture to obtain picture data, and malicious code matching detection is carried out on the picture data through a code feature library; wherein the picture platform is a preset credible picture platform. According to the setting, behavior characteristics of downloading the picture file by the process can be monitored, the picture flow which is not stored in the ground is deeply analyzed, malicious loads hidden in the picture flow are accurately identified, and malicious code interception and whole-process disposal of a propagation path of a trusted platform are realized.

The invention discloses a dumper power assembly redundant suspension protection device, a monitoring system and a graded early warning control method.The dumper power assembly redundant suspension protection device comprises at least one redundant suspension assembly, and the redundant suspension assemblies are correspondingly installed on an engine and / or a gearbox and are arranged close to an engine suspension and / or a gearbox suspension; an original suspension fixing hole position of an engine or a gearbox is not shared, and the protection device is used for providing protection when engine suspension and / or gearbox suspension are / is overloaded or failed; the redundant suspension assembly comprises a redundant supporting leg, a redundant shock absorber, an adjusting gasket and an adjusting washer. The redundant suspension assembly does not intervene in work when the vehicle runs normally, interference to original adjustment parameters of the power assembly suspension is avoided fundamentally, it is ensured that the NVH performance of the vehicle is kept at the original level, and both upgrading practicability and driving and riding comfort are considered.

The invention relates to the technical field of livestock and poultry manure treatment, and discloses a livestock and poultry manure treatment technology mode maturity and applicability evaluation method, which comprises the following steps: step 1, determining an evaluation object and a data acquisition range, selecting a livestock and poultry manure treatment technology mode, and determining the evaluation object and the data acquisition range; collecting technical parameters, equipment operation data, economic cost data and environmental emission data of links of excrement collection, storage, transportation, processing and field returning; and 2, constructing a maturity evaluation system, and constructing a target layer-criterion layer-index layer-sub-index layer four-level evaluation system based on an analytic hierarchy process. The problems that an existing evaluation method is single in dimension, high in subjectivity and difficult in quantitative comparison are solved, comprehensive and objective quantitative evaluation on different technical modes is achieved by constructing a four-dimensional evaluation system and introducing multiple quantitative evaluation methods, the evaluation result is scientific and reliable, and the method is suitable for large-scale popularization and application. And powerful decision support can be provided for screening, optimization and popularization of technical modes.

The application discloses a lithium battery state of health estimation method and system based on physical information and Mamba, which comprises the following steps: firstly, the battery voltage and current data of each cycle of the lithium ion battery in the charging stage are acquired, and pretreatment is performed to form a pretreatment data set; secondly, two features of constant voltage charging time and constant voltage charging capacity are extracted based on the pretreatment data set, two features of the maximum incremental capacity and the incremental capacity variance are extracted based on the ICA curve, and a feature vector is constructed; then, the feature vector is input into a PINN-Mamba neural network to obtain a SOH prediction value; finally, the PINN-Mamba neural network is optimized and trained by using a loss function, and is tested and verified. The application solves the problem of uneven distribution of features of the battery in different states, and effectively improves the prediction accuracy of the SOH.

The application provides a three-dimensional atmospheric tomography method and system for low-altitude airborne Beidou observation, comprising: obtaining dynamic airborne Beidou observation data, processing the dynamic airborne Beidou observation data by using a preset GNSS processing software to obtain an airborne information set; projecting Beidou zenith troposphere delay information onto a slant path by using signal elevation angle and azimuth angle information to obtain Beidou slant path troposphere delay information; determining the boundary space position, horizontal resolution and vertical resolution of each layer of tomography according to the aircraft position information and the Beidou slant path troposphere delay information through a dynamic adaptive tomographic boundary optimization algorithm; constructing a Beidou atmospheric tomography observation equation according to the intercept information of the Beidou signal in the tomographic voxel block; solving the tomographic equation set by using meteorological background field data through an adaptive algebraic reconstruction algorithm; and performing smoothing processing on the tomographic horizontal grid result by using a Gaussian filtering method to obtain the final three-dimensional atmospheric tomographic result.

The application belongs to the technical field of natural language processing and text mining, and discloses a short text theme extraction method fusing a Gaussian mixture model and large model fine tuning. First, short text documents are preprocessed and data is enhanced, and document-level embedding and word-level embedding are obtained by using a pre-training model. Then, a multivariate Gaussian mixture model is constructed to calculate document-theme distribution, and maximum mean difference loss based on Dirichlet prior and document contrast loss are introduced to constrain distribution rationality and enhance semantic consistency. A large language model is used to generate summaries of core words of each theme, and anchor point contrast loss between theme summary vectors and Gaussian theme centers is constructed. The model is fine tuned by combining maximum mean difference loss, document contrast loss and anchor point contrast loss, thereby effectively improving theme consistency of short text theme mining.

The present disclosure provides a pencil beam dose calculation method, comprising: acquiring human body image data; determining a plurality of particle types for treatment; calculating an initial pencil beam dose distribution corresponding to each particle type; determining a target action layer of each dose distribution in the human body image data based on the particle type, wherein the target action layer positions corresponding to different particle types are different; adjusting each initial dose distribution based on the target action layer corresponding to each particle type to obtain a particle pencil beam dose distribution corresponding to each particle type; superimposing the particle pencil beam dose distributions corresponding to each particle type to obtain a target pencil beam dose distribution; wherein the dose proportion of the particle pencil beam dose distribution corresponding to the same particle type in the target action layer is greater than the dose proportion of the particle pencil beam dose of the particle type in other parts of the human body image data.

The application discloses a wheat Tilletia indica-specific biomarker combination, a screening method and application thereof, and the biomarker combination comprises 20-hydroxyarachidonic acid, 1-octadecyl lysophosphatidic acid, linolenic acid, 13-hydroxydocosahexaenoic acid, 4-methoxy estrone, hypoxanthine, 3-hydroxymyristic acid, beta-ecdysterone, 17-phenyltrinorprostaglandin F2 alpha cyclopropylmethylamide, 11,12-dihydroxyeicosatrienoic acid and oxidized octadecadienoic acid. When the biomarker combination is used for identifying Tilletia indica, the time-consuming long and the accuracy easily interfered by the spore development stage limitations of traditional morphological identification are overcome; in view of the molecular level identification difficulty caused by the high genome homology of Tilletia indica and Tilletia controversa, the application digs the feature difference at the level of metabolomics, and constructs a specific chemical fingerprint, so that the technical bottleneck that the close relatives are difficult to accurately distinguish is effectively solved.

The present application belongs to the field of biomedical and cosmetic technology, and particularly relates to the application of protein AMBP in the preparation of a preparation for promoting the activity of hair follicle cells or for preventing and treating alopecia. The present application first discovers that the protein AMBP has the activity of regulating the hair follicle hair papilla cells, hair matrix cells and outer root sheath cells, can promote the proliferation of the hair papilla cells, the hair matrix cells and the outer root sheath cells, and improve the ability of the outer root sheath cells to remove active oxygen; also has the effects of promoting the hair follicle cycle transition, promoting hair growth and hair thickening. Therefore, the protein AMBP can be applied to the preparation of a preparation for treating alopecia.

This invention belongs to the field of integrated wireless and quantum communication technology, specifically an iterative detection method for quantum power domain non-orthogonal multiple access systems. The invention involves constructing a quantum power domain non-orthogonal multiple access system, including a transmitter, a free-space channel, and a receiver. The transmitter sequentially performs channel coding, random interleaving, and quantum coherent state modulation. The receiver includes a Bondurant quantum receiver and a multi-user iterative detector based on the Turbo principle. This invention particularly establishes a photon statistical model at the receiver that considers dead-time effects, incorporating the non-ideal characteristics of the detector into the multi-user detection process. Combined with serial interference cancellation and soft information interaction between the detector and decoder, it achieves iterative detection and decoding of multi-user superimposed signals. This method improves the separation capability, detection accuracy, and decoding reliability of multi-user signals in quantum power domain non-orthogonal multiple access systems.