Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

50results about How to "Improve final performance" patented technology

Domain-adaptive retrieval enhancement generation method and system

The invention discloses a domain-adaptive retrieval enhancement generation method and system, and the method comprises the steps: carrying out the document analysis and semantic partitioning of a document through a semantic understanding model and a document rule, and storing document fragments; on the basis of the large language model and a preset question library, generating a question and disassembling the question; performing correlation evaluation and sorting on the initial retrieval results in combination with a retrieval technology; candidate content most matched with the cue words is screened out through a large language model, and question-retrieval fragments are scored and marked; performing fusion on the initial retrieval result and the scoring and labeling result to construct positive and negative samples, and constructing a supervised data set; and based on a preset loss function, utilizing the supervised data set to carry out supervised training on the knowledge semantic search model, replacing the RAG mixed search module, and carrying out continuous iteration. By means of the technical scheme, new domain knowledge and operation feedback data can be flexibly applied, and domain self-adaptive retrieval enhancement generation is achieved.
Owner:SI-TECH INFORMATION TECH CO LTD

Single-lead electrocardiogram arrhythmia detection method and system based on adaptive attention

The invention discloses a single-lead electrocardiogram arrhythmia detection method and system based on adaptive attention. The method comprises the following steps: preprocessing electrocardiogram data; extracting spatial features and time sequence features of the electrocardiosignals through a multi-scale parallel spatial-temporal feature extraction module; weighting processing is carried out on the time sequence features through an adaptive attention mechanism; splicing and fusing the spatial features and the weighted time sequence features, and carrying out cross-channel information integration to obtain space-time fusion features; the time-space fusion feature outputs a classification result through a classifier comprising stacked BiLSTM and a multi-layer perceptron; and iteratively training and optimizing parameters of the whole network to obtain an optimal classification effect. The system comprises a preprocessing module, a multi-scale parallel spatial-temporal feature extraction module, a weighting module, a splicing fusion module, a classifier module and an iterative optimization module. According to the method, the arrhythmia classification accuracy of the single-lead electrocardiogram data is improved, and misdiagnosis conditions are reduced; the real-time performance of the portable wearable device is improved, and the portable wearable device can be widely applied to multiple fields such as remote medical monitoring, family health management and sports health tracking.
Owner:NANJING UNIV OF SCI & TECH

Construction method of cooperative combat decision-making agent for ground unmanned equipment

The invention relates to a construction method of a cooperative combat decision-making agent for ground unmanned equipment, and aims to solve the problem of space dimension disaster of a complex cooperative combat task. The DRL solution space is effectively compressed, the exploration efficiency is improved, and meanwhile intelligent decision making of tasks with different granularities in the whole combat process is achieved; aiming at the problem of generating a'task layer 'combat command action strategy, a cooperative combat decision-making agent network model is constructed to realize the unified coordination control from the effective representation of complex state input to the combat action; by constructing a distributed training framework, the sampling and training efficiency is improved, and high-throughput asynchronous parallel training of a cooperative combat decision-making agent network model is realized; by establishing a combat decision agent learning algorithm based on deep reinforcement learning and solving the combat decision agent learning algorithm, the learning efficiency and final performance of an agent network model are effectively improved.
Owner:ARMOR ACADEMY OF CHINESE PEOPLES LIBERATION ARMY

AI online education intelligent question and answer information processing method

The invention relates to an AI online education intelligent question and answer information processing method, which comprises the following steps of S1, collecting teaching materials, and performing preprocessing and labeling to form a text data set with labels; s2, constructing a question and answer model based on XLNet; s3, introducing a specific context optimization question and answer model for a specific subject or topic to obtain an optimized question and answer model; s4, constructing an auxiliary understanding information knowledge graph by considering hierarchical design of age groups based on the labeled text data set; s5, inputting a query by the user, and obtaining an optimally matched answer based on the optimized question and answer model; according to the age group of the user and the best matching answer, auxiliary understanding information is obtained from the knowledge graph, and a complete answer is obtained; and S6, adjusting provided content and difficulty through user interaction data, and recommending related domain knowledge based on user behaviors and preferences. According to the method, the direct answers required by the user are provided, the supplementary background information is provided to enhance the understanding, and the requirements of the user are presented in a personalized manner according to the characteristics of the user.
Owner:FUZHOU KLAETTE EDUCATION TECH CO LTD

Hypersonic flight vehicle adaptive attitude control method and system based on near-end strategy optimization

The invention discloses a hypersonic flight vehicle adaptive attitude control method based on near-end strategy optimization, and belongs to the technical field of hypersonic flight vehicle control, and the method comprises the steps: building an air-breathing hypersonic flight vehicle longitudinal dynamic model considering power; performing reinforcement learning training on the obtained environment model by using a PPO algorithm to obtain an optimized control strategy; selecting a proper state representation and action space, and designing a strategy network and a value network for the processed model, so that an original complex attitude control problem is converted into a reinforcement learning problem which can be effectively learned; and training the reinforcement learning model by using an improved PPO algorithm to obtain a final control instruction. According to the method, the model is trained by using the PPO algorithm, so that the defects of complex controller design, strict requirement on model precision and the like of a traditional control method can be effectively overcome. A simulation experiment verifies that pitch angle tracking can be effectively realized, and the stability of an execution mechanism is guaranteed.
Owner:ROCKET FORCE UNIV OF ENG

Method and system for optimizing shape of high-temperature alloy disc type pre-forged piece

The invention belongs to the technical field of material science, and discloses a shape optimization method for a high-temperature alloy disc type pre-forged piece. According to the method, finite element numerical simulation, a support vector machine and a genetic algorithm are combined, a GA-SVR prediction model between an input variable and a target function is established, the optimal value of the target function is predicted, and the optimal geometric parameter combination of the pre-forged piece is obtained. Firstly, the shape of a pre-forged piece is preliminarily designed through experience, then a finite element simulation result is extracted to establish a data set, parameters of a support vector machine model serve as a target function of a genetic algorithm by means of the data set, the parameters of the support vector machine model are optimized by means of the genetic algorithm, and an optimal prediction model is obtained. And for the size parameter combination of the pre-forged piece when the best deformation uniformity evaluation index is obtained, the deformation uniformity of finish forging is effectively improved through global optimization of the genetic algorithm, so that the final performance of the forged piece is improved.
Owner:HUAZHONG UNIV OF SCI & TECH

Die-free mirror image fiber placement method and die-free mirror image fiber placement system

The invention belongs to the technical field of material processing, relates to a mold-free mirror image fiber placement method and a mold-free mirror image fiber placement system, and aims at improving the production flexibility and efficiency by laying fiber ribbons in a mirror symmetry manner without a mold. According to the method, the temperature and pressure of a ribbon are obtained in real time, the output power of a heater and the modified value of the path of a fiber placement tail end movement mechanism are calculated through the temperature difference and the pressure difference, controller parameters are optimized through reinforcement learning, and stable control over the temperature and the pressure is achieved. The system comprises a hardware part and a control system, the hardware part comprises a pressure part, a heat source part and other parts, and the control system ensures the fiber placement quality through real-time feedback adjustment. The problem that the temperature and the pressure are difficult to accurately control in the mold-free mirror image fiber placement process is solved, and the performance of a composite product is improved.
Owner:SHANGHAI THINKHEAD M & E CO LTD

Image magnification method based on dilated convolution fusion

The invention discloses an image magnification method for fusing a multi-stage multi-scale residual structure neural network and a depth separable cavity convolution. The model comprises a shallow image feature extraction network, a deep image feature extraction network and an up-sampling image reconstruction network. The method comprises the following steps: introducing a large kernel attention mechanism, combining three parts of local space convolution, space long-range convolution and channel convolution, and realizing a 21 * 21 receptive field. Secondly, by using the hole convolution, the multi-scale information can be obtained while fewer parameters are used. Wherein the deep information feature extraction network adopts a multi-scale cross network structure, one path is specially used for extracting low-frequency information of an image, and the other path is responsible for extracting high-frequency information. And finally, a complete and clear high-resolution image amplified by four times is obtained by using an up-sampling module.
Owner:ZHEJIANG NORMAL UNIV

Road surface disaster detection method and system, electronic equipment and storage medium

The invention discloses a road surface disaster detection method and system, electronic equipment and a storage medium, and relates to the technical field of intelligent traffic management and computer vision, and the method comprises the steps: collecting image data containing a plurality of road surface disaster types, carrying out the data marking in each image data, and carrying out the data marking; a data sample is constructed through each piece of image data subjected to data annotation; training the improved YOLOv10 model by using the data sample until a loss function of the improved model reaches a preset end condition, and determining the improved model of which the loss function reaches the preset end condition as a disaster detection model; acquiring a to-be-detected image of a to-be-detected road surface, and inputting the to-be-detected image into the disaster detection model for processing to obtain a disaster detection result of the to-be-detected road surface, the disaster detection result including a disaster type and a disaster position; by combining the deep learning technology and optimization algorithm design, the problem that the existing method is insufficient in detection precision, robustness and applicability is solved.
Owner:SICHUAN ENRISING INFORMATION TECH CO LTD

Two-stage sparse method based on small sample fast sparse large language model

The invention discloses a two-stage sparse method based on a small sample rapid sparse large-scale language model, and the method comprises the following steps: S1, mask initialization: based on a plurality of importance indexes for measuring parameters, employing a plurality of mask initialization methods to initialize a mask of a language model; s2, reordering the masks: reordering the masks based on the small sample calibration set and the straight-through estimator; s3, mask adjustment: based on the small sample calibration set and the reordered mask matrix in the step S2, taking an original model as a teacher model, taking a sparse model as a student model, realizing a block-level self-distillation process of the large language model in a time dimension before and after sparsification, and obtaining the sparse model; the method consumes less computing power, uses a smaller data set, occupies less GPU memory while realizing considerable performance, does not influence the reasoning speed of the sparse model, and has remarkable practical value and application prospect.
Owner:XIAMEN UNIV

A heat treatment device for annealing steel pipe

The present invention relates to the technical field of steel pipe production, and specifically to a heat treatment device for annealing steel pipes, comprising a main body, a heating mechanism and a feeding and discharging mechanism, wherein the heating mechanism is arranged on the top of the main body, and the feeding and discharging mechanism is arranged on the top of the heating mechanism, wherein the heating mechanism comprises a heating chamber, wherein the top of the heating chamber is provided with an opening, wherein a rotating frame is rotatably connected inside the heating mechanism, wherein the circumferential outer surface of the rotating frame is provided with an installation groove, wherein a clamping mechanism is arranged inside the installation groove for clamping and rotating the steel pipe to achieve uniform heating, wherein the feeding and discharging mechanism comprises a limiting shell fixedly installed at the top opening of the heating chamber, wherein cavities are arranged inside the two ends of the limiting shell. The present invention realizes continuous self-rotation heating of the steel pipe by combining the rotating frame with the clamping mechanism, wherein the automatic feeding and discharging of the feeding and discharging mechanism shortens the production cycle, wherein the continuous rotation of the rotating frame supports the simultaneous annealing of multiple pipes, thereby meeting large-scale production and reducing labor intensity.
Owner:HUNAN YIGE NEW MATERIAL TECHNOLOGY CO LTD

Blood filtering membrane preparation device and method

The invention discloses a blood filtering membrane preparation device and method, and belongs to the technical field of blood filtering membrane preparation. The method comprises the steps that materials are prepared, a membrane material is polyethersulfone, a solvent is a mixed solvent and specifically comprises N-methyl pyrrolidone (NMP) and gamma-butyrolactone, additives comprise a pore-forming agent, an anticoagulant and bioactive molecules, and the bioactive molecules are heparin. A spinning solution is uniformly dispersed through the ultrasonic dispersion instrument and the magnetic stirrer, the agglomeration state of a membrane material and an additive in a solvent is effectively broken through through the strong vibration and cavitation effect of the ultrasonic dispersion instrument, and full contact and mixing of all components are promoted; therefore, the uniformity of a mixed solution before spinning is remarkably improved, a blood filtering film forming structure is more compact, pore distribution is more uniform, the spinning solution is stirred in cooperation with a magnetic stirrer, agglomeration of particles in the spinning solution is reduced, and the uniformity and stability of the spinning solution can be more effectively improved.
Owner:DALTON BIOTECHNOLOGY (SHENZHEN) CO LTD

Formula of series green and environment-friendly mortar suitable for northwest climate

The invention discloses a series green and environment-friendly mortar formula suitable for northwest climate, and particularly relates to the technical field of mortar, the series green and environment-friendly mortar formula comprises the following components: 100-250 parts of cement, 20-30 parts of attapulgite powder, 20-30 parts of boric sludge, 150-250 parts of natural sand, 10-20 parts of reinforcement powder, 5-10 parts of polysiloxane elastomer, and 2-4 parts of itaconic acid. The prepared mortar is better in breaking strength, corrosion resistance and water resistance, meanwhile, the mortar can be toughened and enhanced, the shrinkage resistance of the mortar is improved, early-stage shrinkage cracking is reduced, then the prepared mortar can effectively slow down the evaporation speed of water, the wettability of the mortar is kept, and the water resistance of the mortar is improved. The mortar prepared by the invention has better weather resistance, can resist strong sunlight, high temperature, low temperature and other extreme weather conditions in northwest regions, keeps the long-term stability and durability of the mortar, and can prevent the mortar from being frozen and damaged under the severe cold weather condition in winter in northwest regions.
Owner:NINGXIA ZHISHANG NEW MATERIAL TECHNOLOGY CO LTD

A wind power conversion efficiency management method

The present invention relates to the technical field of wind power efficiency management, and specifically to a method for managing wind power conversion efficiency, including: determining the daily power consumption, power generation, power supply pressure, and maintenance power consumption of a wind power station under different weather conditions. Determine daily power consumption forecast data, power generation forecast data, and power supply pressure forecast data based on weather forecast data. Determine a maintenance work plan based on weather forecast data, and determine maintenance power consumption forecast data. Determine a "power surplus-time" relationship. When the power surplus is positive, the excess power is intercepted and stored in an electric energy storage module, so that the electric energy stock in the electric energy storage module matches the maintenance power consumption forecast data. When the electric energy stock in the electric energy storage module matches the maintenance power consumption forecast data, the corresponding maintenance work is performed. While ensuring the normal operation of the wind power station, it improves the ultimate efficiency of wind power and avoids unnecessary energy losses.
Owner:SICHUAN SHUWANG TECH +1

Stirring device for chemical raw material production

The invention provides a stirring device for chemical raw material production, and belongs to the technical field of chemical raw material production, the stirring device comprises a stirring barrel and a stirring shaft arranged in the stirring barrel, and the bottom of the stirring barrel communicates with a discharge pipe; the stirring shaft is hollow and rotationally connected to the middle of the stirring barrel, a plurality of stirring blades are arranged on the stirring shaft, a plurality of material guiding openings communicated with an inner cavity of the stirring shaft are formed in the bottom of the stirring shaft, a plurality of filtering holes are formed in the top of the stirring shaft in an array mode, and an auger is rotationally connected to the interior of the stirring shaft; an auxiliary stirring blade capable of rotating along with the auger is arranged between the upper surface of the auger and the top wall of the inner cavity of the stirring shaft; a driving unit is arranged at the bottom of the stirring barrel and is used for driving the stirring shaft and the auger to rotate in opposite directions; according to the invention, the influence of solid impurities on stirring work and defoaming work can be reduced.
Owner:NANYANG DESHIWEI DIGITAL MATERIAL CO LTD

Gastrointestinal tract image classification method based on improved Adam optimization algorithm

The invention discloses a gastrointestinal tract image classification method based on an improved Adam optimization algorithm, and belongs to the technical field of medical image processing. The method comprises the following steps: a data preprocessing module for acquiring gastrointestinal tract image data for preprocessing, constructing a data set, and randomly dividing the preprocessed gastrointestinal tract image into a training set and a test set; the feature extraction module is used for extracting key features from the preprocessed image; the image classification module is used for classifying the images subjected to feature extraction by adopting an improved Adam algorithm (AdaWAGC algorithm); the result output module is used for displaying an image classification result; and the evaluation module is used for evaluating the classification result. By adopting the method to classify the gastrointestinal tract image, the accuracy and efficiency of gastrointestinal tract image recognition can be improved.
Owner:SHENYANG UNIV

A method for removing micro-water from polar lubricating oil based on coordination of electric field polarization and lattice field

PendingCN122503166Aavoid damageWeaken hydrogen bond associations
This invention discloses a method for removing micro-water from polar lubricating oil based on the synergy of electric field polarization and crystal field. The method includes: first, filtering the polar lubricating oil containing trace amounts of water to remove mechanical impurities, then heating it to 25℃~35℃; then introducing it into an electric field polarization unit, where it is polarized using an electric field until the dielectric constant of the lubricating oil increases by 3%~8% relative to its initial value; next, introducing it into a crystal field unit, where it is stirred and mixed to fully contact an inorganic crystalline particle adsorbent with a superhydrophilic surface, using the adsorbent to remove water from the oil; finally, passing it through a fine filtration unit to remove the water-swollen inorganic crystalline particle adsorbent, thus completing the micro-water removal of the polar lubricating oil. This invention achieves efficient and deep dehydration of polar lubricating oil at room temperature, with advantages such as low energy consumption, low cost, no additive loss, and no impact on oil performance.
Owner:CHONGQING TECH & BUSINESS UNIV

Complex special-shaped section launching barrel forming process

The invention provides a forming process of a complex special-shaped section launching barrel. The forming process comprises the following steps of production preparation, forming of a functional layer and an inner pipe body, filling of a guide sliding groove and forming of an outer pipe body, overall curing and pre-pipe-hoop forming and aftertreatment. According to the invention, a sharp and complex geometrical shape is converted into a curved surface in smooth transition through the pretreatment procedure of filling wires into the guide chute, so that a perfect base is provided for the subsequent laying of a structural layer, and the hidden danger of overhead is fundamentally eliminated; a vacuum bag technology is combined with hand lay-up forming, and a silicone rubber block is used in a key area for local sealing and flow guiding. According to the combined process, uniform compaction pressure can be applied within the range of the whole structure, interlayer bubbles and redundant resin are efficiently removed, a product with high compactness and low porosity is obtained, and the reliability of consistency of mechanical properties of the product is greatly improved.
Owner:ANHUI YINGLIU HAIYUAN COMPOSITE MATERIALS TECH CO LTD

Fan blade damage identification method and device based on small sample

The invention belongs to the field of neural networks, and particularly discloses a fan blade damage identification method and device based on a small sample, and the method comprises the steps: collecting an audio time domain signal of a fan blade, and converting the audio time domain signal into a frequency domain signal; performing feature extraction on the frequency domain signal to obtain a dimensionless octave feature, and taking the octave feature as a data set for model training; and inputting the data set into an initial neural network model, carrying out gradient updating on the initial neural network model by using an MAML algorithm, and continuously adjusting parameters of the initial neural network model until the model converges to obtain a trained fan blade damage identification model, and performing fan blade damage identification by using the trained fan blade damage identification model to obtain an identification result of the fan blade. According to the invention, the model training efficiency and training precision can be improved, and the recognition accuracy of the fan blade is improved.
Owner:HUAZHONG UNIV OF SCI & TECH RES INST SHENZHEN

Flexible transparent epoxy resin as well as preparation method and application thereof

The invention discloses flexible transparent epoxy resin as well as a preparation method and application thereof.The preparation method comprises the following steps: firstly, adding an amino silane coupling agent into an epoxy resin monomer at normal temperature, and uniformly stirring to obtain a mixed solution; then heating the mixed solution for reaction, and after the reaction is finished, carrying out vacuum treatment on the reaction solution to obtain a precursor solution; and finally, carrying out curing treatment on the precursor solution to prepare the flexible transparent epoxy resin. The preparation method is simple in process and reasonable in design, and the obtained flexible epoxy resin has good mechanical strength, thermal stability and transparency.
Owner:SHAANXI UNIV OF SCI & TECH

A gear steel for wind turbine

The present invention provides a gear steel for a wind turbine. The gear steel comprises a gear steel body and groove steps symmetrically arranged at both ends, and the bottom and upper edges of the groove steps have transition fillets with a radius of 1 / 8 of the width of the groove steps. This design significantly reduces stress concentration at sharp corners, reduces the risk of cracks and fractures, and improves the fatigue strength and durability of the gear steel. The forging process includes preheating, preliminary forging, remelting and heat preservation, mold forming, cooling, post-processing and surface treatment. By multiple step-by-step forging and controlling the deformation amount each time, local stress can be effectively dispersed and released, and the material fluidity and mechanical properties can be optimized, thereby significantly improving the overall quality, stability and service life of the gear steel.
Owner:ZHEJIANG JIEDE MASCH TECH CO LTD

Artificial plastic stone design and construction method based on artificial intelligence technology

The invention discloses an artificial plastic stone design and construction method based on an artificial intelligence technology, the method runs through the whole process of design and construction through an AI technology, and the method comprises the following specific steps: firstly, writing cue words, and generating a large number of concept and intention diagrams of artificial plastic stones by using AI; then, through an image fusion technology, a plurality of design schemes are generated in combination with classical design case optimization, and after an optimal scheme is selected, a corresponding three-dimensional model is generated by utilizing AI; generating a structure model of the supporting steel frame, the modeling reinforcing mesh and the concrete lining mesh through the three-dimensional model; and in the construction stage, reinforcing steel bar machining, concrete pouring and detail carving are conducted according to the model, and finally the accurate plastic stone structure is formed. According to the method, the AI technology is integrated into multiple working processes of the landscape garden, intelligence from conceptual design to the whole construction life cycle is achieved, the design efficiency is greatly improved, the problem that design and construction are disjointed in a traditional method is solved, and the method has the advantages of being efficient, accurate and sustainable.
Owner:SHANGHAI GARDENS (GROUP) CO

An unsupervised exposure correction method

PendingCN122597232Arestore brightnessImprove reliability
The application discloses a kind of unsupervised exposure correction methods, method includes: constructing semantic prompt guide's exposure perception module, the exposure level of image is perceived at visual semantic level;Exposure parameter modulation is constructed pseudo-supervision generation module, according to image exposure level adaptive exposure modulation is carried out, and synthesis is used to guide model training pseudo-supervision image;Construct color degradation compensation exposure correction module, by joint execution luminance correction and color compensation, the luminance distortion in abnormal exposure image and color degradation are recovered;Unsupervised exposure correction network is trained using the network optimization mechanism of multiple loss functions;Based on the unsupervised exposure correction network after training, input image is handled, and image luminance level and color fidelity are effectively improved.The application generates pseudo-supervision image according to image exposure level adaptive to guide model training, and compensates the color information of image while luminance correction, to realize the collaborative recovery of luminance and color.
Owner:TIANJIN UNIV

Cosmetic

Provided is a cosmetic comprising composite particles comprising an inorganic nucleating agent which is at least one selected from the group consisting of glass particles and poorly soluble sulfate particles and a metallic soap which is a salt of a fatty acid and at least one selected from the group consisting of calcium and magnesium; with this cosmetic, the adhesion after application and / or the final effect after application are also excellent in addition to the spreadability during application, and the cosmetic has a good feeling of use.
Owner:IWASE COSFA CO LTD

A boiler circulating water treatment device and process

The present application provides a boiler circulating water treatment device and process, belonging to the field of water treatment technology, comprising a treatment box and an anode rod, a cathode rod and an injection assembly arranged therein, wherein a water outlet pipe is connected to the box wall of the treatment box; the injection assembly comprises a receiving plate and a supporting ring which are rotatably connected to the left and right inner walls of the treatment box respectively, a plurality of injection pipes are arranged in a circumferential array between the receiving plate and the supporting ring, and a water supply unit is arranged between the supporting ring and the treatment box; a plurality of guide blades are rotatably sleeved on the outside of the injection pipes, a guide groove is opened on the inner wall of the treatment box, a plurality of guide columns are slidably connected inside the guide groove; a driving member is arranged on the side wall of the treatment box; the present application can reduce the probability of surface wear of the cathode rod, and can also improve the fluidity of the boiler circulating water inside the treatment box to a certain extent.
Owner:GUANGZHOU BAIHONG NEW ENERGY TECH CO LTD

Boiler circulating water treatment device and process

The invention provides a boiler circulating water treatment device and process, and belongs to the technical field of water treatment.The boiler circulating water treatment device comprises a treatment box, an anode bar, a cathode bar and a spraying assembly, and the anode bar, the cathode bar and the spraying assembly are arranged in the treatment box; the spraying assembly comprises a bearing disc and a bearing ring which are rotationally connected to the left inner wall and the right inner wall of the treatment box respectively, a plurality of spraying pipes are arranged between the bearing disc and the bearing ring in a circumferential array mode, and a water supply unit is arranged between the bearing ring and the treatment box; a plurality of flow guide fan blades are rotationally arranged outside the injection pipes in a sleeving mode, a guide groove is formed in the inner wall of the treatment box, and a plurality of guide columns are slidably connected into the guide groove; a driving part is arranged on the side wall of the treatment box; the probability of surface abrasion of the cathode bar can be reduced, and the flowability of boiler circulating water in the treatment box can be improved to a certain extent.
Owner:GUANGZHOU BAIHONG NEW ENERGY TECH CO LTD

Language model training method and training device

The invention provides a language model training method and device, and the method comprises the steps: obtaining original dialogue type text data, and converting dialogue type data in the original dialogue type text data, so as to obtain a plurality of pieces of text type data; wherein the original dialogue type text data comprises at least two types of text data; performing text data mixing on the multiple pieces of text type data in a preset mixing mode to obtain target training data; wherein the preset mixing mode comprises a title-free mixing mode and / or a multi-title mixing mode; and inputting the target training data into a to-be-trained language model, and training the to-be-trained language model by using the target training data to obtain a target language model. Through the method and the device, the calculation overhead of the model is controlled while the text expression diversity is ensured, so that the training efficiency and the final performance of the language model are remarkably improved.
Owner:SHANGHAI XIYU JIZHI TECH CO LTD

Question answer generation method and device, equipment and medium

The invention relates to the technical field of intelligent decision making, can be applied to business system platforms such as financial science and technology, medical health and the like, and discloses a question answer generation method, device, equipment and medium, and the method comprises the steps: collecting a plurality of question answers generated by questions, collecting the question answers generated by each question into a question and answer sample set, calculating a reward model score and a reference model benchmark score of each question and answer sample in the question and answer sample set one by one, fitting into a confidence coefficient curve, and dividing an answer retention interval; screening answers to the questions according to the answer retention interval, and calculating and updating an answer retention proportion in the question and answer sample set; optimizing the question and answer model according to the answer retention proportion, and generating a plurality of candidate answers of the user question by using the optimized question and answer model; and calculating candidate reward scores, retaining candidate answers corresponding to the candidate reward scores in the answer retaining interval, and performing semantic fusion on the answers to generate a final question answer of the user question. The accuracy of question answer generation can be improved.
Owner:PING AN TECH (SHENZHEN) CO LTD

Tire building machine quality defect named entity identification method based on improved BiLSTM-CRF

PendingCN119623467AImprove unstructured dataDispatch in timeSemantic analysisText processingData setNamed-entity recognition
The invention belongs to the field of named entity recognition of natural language processing, and discloses a tire building machine quality defect named entity recognition method for improving BiLSTM-CRF. Comprising the following steps: step 1, preprocessing unstructured data such as on-site production quality of a forming machine, determining a preset BME label, performing manual labeling by using a labeling tool, and providing a data set for a subsequent process; and step 2, constructing a BiLSTM-CRF training learning model. And step 3, adding a learning rate scheduler to carry out timely dynamic adjustment on model parameters. And 4, testing the trained model by using a test set. Test results of the HMM model, the BiLSTM model, the BiLSTM + CRF model and the improved BiLSTM + CRF model are compared, and the results show that compared with other models, the improved BiLSTM-CRF model can consider context information and can dynamically adjust parameters, so that the model can more accurately judge entity boundaries and categories, and the overall effect is improved compared with other models.
Owner:QINGDAO UNIV OF SCI & TECH

Elastic skin-friendly heat-insulating fire-resistant sensing yarn and spinning method and device thereof

The invention provides an elastic skin-friendly heat-insulation fire-resistant sensing yarn and a spinning method and device thereof.The yarn is obtained by installing a wrapping spinning device on a ring spinning frame for spinning, and the yarn is of a multi-layer structure; the elastic core yarn, the heat-insulating inorganic fiber wrapping layer, the conductive fiber wrapping layer and the flame-retardant short fiber wrapping layer are sequentially arranged from inside to outside; the core yarn rotates under the twisting effect, and then the heat-insulating inorganic fiber filaments, the conductive fiber filaments and the flame-retardant short fiber strands are driven to revolve around the core yarn for wrapping; due to the fact that the elastic fibers are adopted as the core yarn, the elastic fibers have the good tensile property, meanwhile, the heat-insulation inorganic fiber filaments and the conductive fiber filaments can provide the heat-insulation, flame-retardant and conductive functions, and the flame-retardant short fibers on the outermost layer have the skin-friendly and flame-retardant effects. The elastic high-temperature-resistant sensing yarn solves the problem that elastic high-temperature-resistant sensing yarn lacks at present, and the elastic high-temperature-resistant sensing yarn can be applied to the fields of fire rescue, aerospace military industry and the like in the future.
Owner:WUHAN TEXTILE UNIV