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

152results about How to "Clear boundaries" patented technology

Advanced nasopharynx cancer treatment effect prediction system based on deep learning

The invention relates to the technical field of medical data processing, in particular to an advanced nasopharynx cancer treatment effect prediction system based on deep learning. According to the invention, a data acquisition module acquires multi-modal medical images and clinical data of a patient from a hospital information system; the data preprocessing module is used for cleaning, registering, normalizing and standardizing the data; the feature extraction module extracts image features by using a convolutional neural network and fuses the image features with clinical features to form a comprehensive feature vector; the model construction and training module constructs a deep learning model, and is adjusted and optimized through training; the treatment prediction module inputs the comprehensive feature vector to obtain a treatment effect prediction result; and the result evaluation and feedback module compares prediction and actual results, calculates indexes and collects and feeds back the improved model. The system comprehensively integrates data, accurately mines features, constructs an effective model, provides personalized prediction to assist clinical decision, continuously improves the model performance, and has important significance in improving the prediction accuracy of the treatment effect of advanced nasopharynx cancer, promoting personalized medical treatment and the like.
Owner:THE FIRST AFFILIATED HOSPITAL OF FUJIAN MEDICAL UNIV

Intelligent factory logistics automatic scheduling system and method thereof

An intelligent factory logistics automatic scheduling system comprises a data acquisition module, a scheduling optimization module, a plan execution module, a monitoring early warning module and a system management module. Wherein the data acquisition module is used for acquiring real-time data of workshop production progress, warehouse stock and material distribution, and performing data preprocessing and storage management; the scheduling optimization module is used for establishing a multi-target scheduling optimization model based on a machine learning algorithm and dynamically generating an optimal scheduling decision; the plan execution module is used for converting the scheduling plan into an instruction sequence which can be executed by logistics equipment so as to realize automatic distribution and carrying of materials; the monitoring and early warning module is used for tracking a logistics execution state and carrying out early warning and adjustment on abnormal conditions; and the system management module provides task management, parameter configuration and user management system functions, guarantees stable operation of the system, and constructs a data-driven intelligent logistics scheduling system by applying frontier artificial intelligence technologies such as multi-agent reinforcement learning, operation optimization, transfer learning and the like.
Owner:GUANGDONG JINGGONG INTELLIGENT SYST CO LTD

Road user track and behavior prediction method in dense heterogeneous traffic environment

The invention discloses a road user track and behavior prediction method in a dense heterogeneous traffic environment, and the method comprises the steps: carrying out the target detection and recognition of each road user in a video image sequence frame based on a deep convolutional neural network, and taking a recognition result as a road user type constraint; taking the detected and identified track coordinates and timestamp information of the road user as historical track information of the road user; according to the method, on one hand, a traffic map sequence of a target road user per second within a tf time period after a timestamp t is obtained through prediction, the traffic map sequence is subjected to spectral clustering and then serves as a map regularization constraint in a regularization loss function form to be used for correcting future trajectory coordinates of the road target user output by semi-supervised pipeline prediction; and on the other hand, entering a logistic regression function network after passing through a full-connection network, and outputting the behavior prediction type of the road user. According to the method, the track and behavior of the road user can be accurately predicted in a dense heterogeneous traffic environment.
Owner:NANTONG UNIV

Training method and system based on circuit board defect detection model

The invention relates to the technical field of image processing, in particular to a training method and system based on a circuit board defect detection model. The method comprises the following steps: acquiring an automatic optical monitoring image of the circuit board, and carrying out circuit board surface defect detection on the automatic optical monitoring image of the circuit board so as to obtain surface defect data of the circuit board; marking a circuit board defect area according to the circuit board surface defect data so as to obtain circuit board defect area data; carrying out welding spot three-dimensional contour imaging according to the automatic optical monitoring image of the circuit board so as to obtain welding spot three-dimensional contour data, and carrying out circuit board pseudo soldering analysis according to the welding spot three-dimensional contour data so as to obtain circuit board pseudo soldering data; and performing electrical connection defect analysis according to the circuit board defect area data so as to obtain electrical connection defect data. Based on the computer vision technology, the accuracy of the circuit board defect detection model is improved.
Owner:SHENZHEN SANDEYING ELECTRONICS CO LTD

Mine transient electromagnetic three-component detection method

The invention relates to the technical field of mine electromagnetic detection, and discloses a mine transient electromagnetic three-component detection method, which comprises the following steps: S1, selecting and preparing equipment, selecting transmitting equipment, selecting receiving equipment which comprises three sets of receiving coils or antennas consistent with the measurement direction of a three-component electromagnetic field, and configuring data acquisition equipment with anti-interference capability; s2, laying and planning measuring points, dividing an exploration area according to geological data, and dividing the area into a plurality of measuring grids; s3, calibration and system inspection: carrying out system calibration on the transmitting and receiving equipment; s4, formal data acquisition; s5, data storage and backup; s6, data preprocessing; s7, performing inversion and modeling; s8, performing three-dimensional imaging and geological interpretation; and S9, compiling a report. The three-component data of the electromagnetic field is collected, the spatial change characteristics of the electromagnetic field are comprehensively captured, the position, shape and conductivity distribution of the underground anomalous body are detected more accurately, and the limitation caused by traditional single-component measurement can be effectively solved.
Owner:HUANENG COAL TECH RES CO LTD +2

Entity gradient dot matrix mixed structure multi-scale topological optimization method

The invention provides a multi-scale topological optimization method for an entity gradient dot matrix mixed structure, and belongs to the field of structural topological optimization design, and the off-line stage comprises the following steps: constructing parameterized entity microstructure and dot matrix microstructure unit cells, establishing a sample database and an agent model, and realizing rapid mapping of unit density and unit cell mechanical attributes; the online stage comprises the following steps: defining double design variables of entity unit density and dot matrix unit density, minimizing structural flexibility as a multi-scale topological optimization model of a target function, introducing a Heaviside function to project the entity unit density, completing model reconstruction of the unit density and a corresponding entity gradient dot matrix structure, and obtaining a multi-scale topological optimization model of the target function. And obtaining a mixed structure topological configuration in which a solid structure and a gradient lattice structure are reasonably distributed. According to the method, co-parallel topological optimization of the entity structure and the gradient lattice structure is achieved, the comprehensive mechanical property of the structure is remarkably improved, and a new method is provided for entity gradient lattice mixed configuration design.
Owner:ANHUI POLYTECHNIC UNIV MECHANICAL & ELECTRICAL COLLEGE

Low-temperature sintered asymmetric porous silicon carbide ceramic membrane separation layer and preparation method thereof

The invention discloses a low-temperature sintering asymmetric porous silicon carbide ceramic membrane separation layer and a preparation method thereof, the preparation method comprises the following steps: a) preparation of a coating liquid: firstly dispersing a binder in a solution, adding silicon carbide particle powder, carrying out magnetic stirring dispersion for a period of time, then adding a dispersant, and stirring; then putting the stirred and mixed film coating liquid and the grinding balls into a ball milling tank for ball milling; and adjusting the pH value of the ball-milled silicon carbide coating liquid with ammonia water, and carrying out vacuum defoaming to finally obtain the high-stability and uniformly-dispersed silicon carbide coating liquid. Wherein the addition amount of the silicon carbide coarse particles is 0.5-50 wt% of the mass of silicon carbide; b) coating; c) drying; and d) sintering. The asymmetric porous silicon carbide ceramic membrane prepared by the method is low in sintering temperature, and the prepared separation layer membrane is controllable in pore size and uniform in pore size distribution; the method has excellent application prospects in the field of membrane separation.
Owner:NANJING TECH MEMBRANE APPLICATION TECHNOLOGY RESEARCH INSTITUTE CO LTD +2

Focused ion beam ultrathin section sample preparation pretreatment method for wafer sample

The invention relates to a focused ion beam ultrathin section sample preparation pretreatment method for a wafer sample, and belongs to the technical field of wafer material microscopic experiments. The wafer sample preparation pretreatment method provided by the invention comprises the following steps: (1) depositing a metal oxide on the surface of a wafer sample by adopting an atomic layer deposition technology to obtain an ALD layer; (2) coating a curing glue solution on the surface of the ALD layer, and curing to obtain a curing glue layer; and (3) plating a conductive protection layer on the surface of the cured adhesive layer to finish sample preparation pretreatment. The sample preparation pretreatment method disclosed by the invention can well solve the problems of irradiation damage and deformation of a fine structure which is not resistant to electron beam irradiation, curtain effect in a cutting process, unobvious structural boundary in a shooting process and the like in focused ion beam ultrathin section sample preparation, and has good application prospects in the technical field of wafer material microscopic experiments. And the practical value is very high.
Owner:GUANGZHOU GRG METROLOGY & TEST CO LTD WUXI

Method for processing silicon carbide positioning edge, silicon carbide crystal with positioning edge and application of silicon carbide crystal

The invention discloses a method for machining a silicon carbide positioning edge, a silicon carbide crystal with the positioning edge and application, and belongs to the technical field of silicon carbide machining. The method comprises the following steps: (1) fixing a silicon carbide crystal; (2) orienting the crystal face (0004) or the crystal face (11-28) of the silicon carbide crystal by adopting XRD (X-Ray Diffraction), firstly adjusting an incident angle theta 1 and an emergent included angle theta 2 of an X-ray, then rotating the silicon carbide crystal by taking a vertical center line as an axis and performing phi scanning until the silicon carbide crystal rotates to a position with the maximum diffraction intensity, then rotating the silicon carbide crystal by taking a horizontal center line as an axis and performing x scanning, and finally obtaining the crystal face (0004) or the crystal face (11-28) of the silicon carbide crystal. And finally, continuously rotating the silicon carbide crystal by taking the horizontal center line as the axis, and carrying out chi value adjustment according to the required crystal drift angle. And (3) performing positioning edge marking on the (1-100) crystal face of the crystal by adopting micro jet water-guide laser to form a positioning edge. According to the machining mode, material waste can be avoided, and the machining efficiency and machining precision of the positioning edge are improved.
Owner:SICC SHANGHAI CO LTD

Photovoltaic station fine identification method and device based on classifier and deep learning

The invention belongs to the technical field of remote sensing detection, and particularly discloses a photovoltaic station fine identification method and device based on a classifier and deep learning, and the method comprises the steps: carrying out the preprocessing of a multi-source remote sensing image of a to-be-identified photovoltaic station in a research region range; inputting the preprocessed multi-source remote sensing image into a trained machine learning classifier to carry out preliminary coarse resolution recognition, and obtaining a high-resolution remote sensing image containing a coarse recognition contour of a potential photovoltaic region; and inputting the high-resolution remote sensing image containing the coarse identification contour of the potential photovoltaic region into a trained deep learning model for fine identification, and outputting a segmentation mask with panel type information, the segmentation mask being used for distinguishing a photovoltaic panel and a panel gap. According to the method, the potential area of the photovoltaic station can be quickly positioned in the specified area range, and fine identification is carried out on the basis, so that the accuracy and efficiency of identifying the photovoltaic station in a complex and diversified area are improved.
Owner:HUAZHONG UNIV OF SCI & TECH

Microtopography recognition method, device and equipment and storage medium

The invention provides a microtopography recognition method, device and equipment and a storage medium, and the method comprises the steps: searching an optimal calculation window of a topography position index based on a mean change point analysis method, and obtaining the topography position index features of a target sample based on the optimal calculation window; obtaining the size of a neighborhood window when the average elevation deviation reaches a preset threshold value based on the average elevation deviation counted in the neighborhood window of the elevation points in the target sample, and calculating a section curvature feature based on the obtained size of the neighborhood window; constructing a sample feature data set at least based on the terrain position index features and the section curvature features; training a microtopography recognition model based on the sample feature data; and inputting the elevation data of the to-be-detected area into the trained microtopography recognition model to recognize the microtopography of the to-be-detected area. The optimized characteristic factors and the combination thereof can realize effective distinguishing of the microtopography, so that boundaries of various microtopographies are clearer, and the microtopographies can be quickly identified.
Owner:WUHAN UNIV

Logistics policy document analysis method based on multi-modal features

The invention provides a logistics policy file analysis method based on multi-modal features, which comprises the following steps of: extracting key words of themes from a logistics policy file, and performing importance score sorting on the key words to obtain the sorted key words, so as to clarify the boundary of the key words; furthermore, by means of a timestamp marking mode, multi-modal keywords fusing time sequence features are obtained, a more accurate prediction effect is achieved, and a foundation is further laid for subsequent evaluation of logistics policy documents according to time periods. In addition, similarity evaluation is carried out on the superior logistics policy file and the subordinate logistics policy file through a similarity function, and the superior and subordinate collaboration degree is obtained to evaluate the implementation efficiency of the logistics policy file. In addition, in the process of obtaining the superior policy document, a more accurate keyword of the theme is obtained by adjusting the preset weight of the theme at the subtitle.
Owner:EAST CHINA JIAOTONG UNIVERSITY

Remote sensing ground feature classification extraction network and drawing method in complex scene

The invention relates to the technical field of remote sensing image processing, and discloses a remote sensing ground feature classification extraction network and drawing method in a complex scene, and the method comprises the following steps: S1, obtaining high-resolution satellite remote sensing image data, carrying out the preprocessing of an image, and generating a multi-scene ground feature data set; s2, constructing a two-layer U-shaped network comprising an encoder and a decoder, wherein the two-layer U-shaped network is used for multi-scale feature extraction and ground feature boundary segmentation; s3, inputting the preprocessed remote sensing image into a two-layer U-shaped network, and carrying out ground feature classification and segmentation; s4, performing splicing processing on the segmentation result, and generating a ground feature type spatial distribution map; and S5, performing noise elimination and post-processing operation on the generated distribution map. The multi-scale features are selectively extracted through the feature selection module (FSM), the attention of the network to small target ground features can be enhanced, and the segmentation precision of the small-scale ground features is remarkably improved by selecting the features containing important details.
Owner:HENAN AGRICULTURAL UNIVERSITY

Shale pore type identification and quantitative characterization method based on digital core

The invention provides a shale pore type identification method based on a digital core, and the method comprises the steps: carrying out the gray adjustment processing and image noise reduction processing of an original gray image of a to-be-identified shale sample, and obtaining a standard gray image; organic matter and transition zone images in the standard grayscale image are extracted, organic matter identification processing is carried out, and an organic matter connected body image is determined; extracting a pore image in the standard gray-scale image, performing pore recognition processing in combination with the organic matter connected body image, and determining an organic pore image and an inorganic pore image; and performing micro-crack identification processing on the inorganic pore image, and determining a micro-crack image and an inorganic pore image. By applying the method, the organic pores, the inorganic pores and the microcracks of the shale can be quantitatively identified, the pore structure of the shale can be quantitatively represented, and the structural parameters of the organic pores, the inorganic pores and the microcracks can be analyzed.
Owner:CHINA PETROLEUM & CHEMICAL CORP +1

Electric pulse nanometer knife applied to intestinal cancer ablation

The invention discloses an electric pulse nanometer knife applied to intestinal cancer ablation. The electric pulse nanometer knife comprises an elastic tube body, an elastic balloon and a flexible electric insulation covering film covering the outer wall of the elastic tube body. The inner wall of the elastic tube body is provided with regularly arranged recesses; the elastic balloon is positioned in the internal space of the elastic tube body; the elastic balloon is provided with an inlet / outlet; flexible electrodes which are regularly arranged are arranged on the surface of the flexible electric insulation covering film; the positions of the flexible electrodes are in one-to-one correspondence with the positions of the recesses; and the flexible electrode is used for introducing electric pulse from the outside to realize IRE treatment. According to the invention, noninvasive intestinal cancer treatment with more accurate positioning and more accurate treatment is realized.
Owner:XIDIAN UNIV

Winnowing and removing device for wizened snakegourd seeds and removing process method of wizened snakegourd seeds

The invention relates to the technical field of winnowing and removing devices, in particular to a wizened snakegourd seed winnowing and removing device which comprises a processing platform, a winnowing and removing mechanism used for separating wizened snakegourd seeds is arranged on the processing platform, and a batch feeding mechanism is arranged in the winnowing and removing mechanism. A winnowing auxiliary mechanism for separating the separated wizened snakegourd seeds is arranged below the winnowing removing mechanism, a rapid discharging mechanism is arranged on the processing platform, the winnowing removing mechanism comprises an air conveying pipeline fixedly installed on the processing platform, and the upper end of the air conveying pipeline is communicated with an air injection disc. According to the snakegourd seed winnowing and removing device, the winnowing and removing mechanism is arranged, the snakegourd seeds can evenly fall down in a rotating mode through mutual cooperation between the rotary feeding assembly and the material guiding inclined plate, the snakegourd seeds can be effectively prevented from being mutually shielded and overlapped in the falling process, and therefore the snakegourd seed winnowing and screening effect is improved; and local accumulation of the snakegourd seeds on the processing platform can be avoided.
Owner:QIANSHAN YOUYU TRICHOSANTHES KIRILOWII MAXIM DEV CO LTD

Sample analyzer, sample analysis method and sample analysis system

The present application relates to a sample analyzer, a sample analysis method and a sample analysis system, including: obtaining the original data corresponding to the sample, the original data including at least one of the original light signal data and the original scatter plot; the original light signal data including at least two paths of light signal data with different angles; inputting the original data into a conversion model, the conversion model including N conversion layers connected in sequence, each conversion layer being used to convert the light signal data according to the corresponding conversion function to obtain converted light signal data; wherein the input of the first conversion layer of the conversion model is the original light signal data, and the input of the Mth conversion layer of the conversion model is the light signal data converted by the M-1th conversion layer; N is greater than or equal to 2, and the maximum value of M is N; generating a converted scatter plot based on the converted light signal data of the Nth conversion layer of the conversion model, and identifying the scatter plot to obtain a particle classification result. This method can improve the accuracy of particle classification.
Owner:SHENZHEN DYMIND BIOTECH

Hydroxyapatite composite polyester microsphere as well as preparation method and application thereof

The invention provides hydroxyapatite composite polyester microspheres as well as a preparation method and application thereof, and relates to the technical field of biomedical materials, the microspheres comprise the following raw materials in parts by mass: 25-95 parts of bioabsorbable polyester and 5-18 parts of hydroxyapatite; the preparation method of the microspheres comprises the following steps: dissolving the bio-absorbable polyester in an organic solvent to obtain a solution 1; adding hydroxyapatite into the solution 1 to obtain a solution 2; dissolving an emulsifier to obtain a solution 3; dropwise adding the solution 2 into the solution 3 to obtain an emulsified solution; cleaning and drying to obtain hydroxyapatite composite polyester microspheres; the microspheres can be used for preparing face filling and wrinkle removing reagents. The microspheres are appropriate in size and clear in boundary, can reduce the possibility of adverse symptoms such as inflammation after skin filling, and are stable and long-acting. The microspheres prepared by the preparation method disclosed by the invention are proper in size, clear in boundary and good in supporting effect after being filled into skin.
Owner:JIAXING RUIQING MEDICAL TECH CO LTD

Dialogue intention analysis method and device, equipment, storage medium and program product

The invention discloses a dialogue intention analysis method and device, equipment, a storage medium and a program product, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining at least one dialogue statement sequence, and carrying out the intention classification analysis of dialogue statements in the at least one dialogue statement sequence, obtaining each type of intention and at least one dialogue statement classified into each type of intention; based on the time sequence of each dialogue statement in the corresponding dialogue statement sequence and the class intention corresponding to each dialogue statement, determining a jump relationship between different class intentions; according to the jump relation between the different types of intentions, an intention relation graph is constructed, the intention relation graph comprises a plurality of nodes and a plurality of directed edges, the nodes represent the types of intentions, and the directed edges represent the jump relation between the different types of intentions; and generating an intention analysis result based on the intention relation graph. The dialogue intention analysis efficiency can be improved.
Owner:WEBANK (CHINA)

Image processing method, model training method and electronic equipment

The invention discloses an image processing method, a model training method and electronic equipment, and relates to the field of large model technology and image processing. The method comprises the steps that an image to be processed is acquired, and the image to be processed comprises a target object; performing feature extraction on the to-be-processed image by using a feature extraction model to obtain article features of the target article; and based on the article features and the edge information of the target article, guiding an image generation model to carry out conversion processing on the to-be-processed image to obtain a target image, other regions except the region where the target article is located in the target image being displayed through a preset texture, and the target article in the target image being in a preset state. The technical problem that the image processing effect is poor in the prior art is solved.
Owner:TAOBAO CHINA SOFTWARE

Rotary batch epitaxy equipment and preparation method thereof

The invention discloses rotary batch epitaxy equipment, which comprises a wafer loading system, an equipment front-end module, a wafer left vacuum chamber, a wafer right vacuum chamber, a wafer transfer chamber, a plasma etching cavity, a heating volatilization cavity, a right reaction module and a left reaction module, and is characterized in that the wafer loading system is used for placing wafer boxes, and 25 wafers are loaded in each wafer box; a manipulator in the equipment front-end module takes out wafers in a wafer box on the wafer loading system, carries the wafers to a wafer left vacuum chamber and a wafer right vacuum chamber after calibration and positioning, a manipulator in the wafer transfer chamber can grab the wafers in the wafer vacuum chambers, loads the single wafers and conveys the single wafers to a subsequent reaction chamber, and the wafer transfer chamber can transfer the single wafers to the subsequent reaction chamber. Selective epitaxy is carried out firstly, then full epitaxy is carried out, a, adhesion and fracture between a wafer boat and a wafer are avoided, and b, the epitaxy temperature is reduced, the growth time is shortened, the boundary of Si and SiGe layers is clear, and the electrical performance of the chip is improved.
Owner:HANGZHOU XINGYUANCHI SEMICON CO LTD

KASP molecular marker related to tomato fruit weight and application of KASP molecular marker

The invention discloses a KASP molecular marker related to tomato fruit weight and application of the KASP molecular marker, and belongs to the technical field of nucleic acid-containing determination methods. The SNP site related to the tomato fruit weight is obtained by using a whole genome association analysis method and is the 336003 site of the total length sequence of the No.9 chromosome of the tomato genome, the polymorphism that T is greater than C exists, and a related KASP molecular marker and a KASP primer are further developed for typing the tomato fruit weight. It is verified that a group typing effect which is highly consistent and clear in boundary is obtained in a tomato group. In the breeding process, allelic variation of the fruit weight is judged according to the genotype, and an excellent population meeting the breeding target size can be quickly selected in the seedling stage, so that early screening of the breeding population is realized, and the molecular breeding process of tomato varieties is accelerated.
Owner:ZHEJIANG ACADEMY OF AGRICULTURE SCIENCES

Monodisperse vaterite-shaped calcium carbonate with uniform particle size and preparation method thereof

The invention discloses a preparation method of monodisperse vaterite-shaped calcium carbonate with uniform particle size, which comprises the following steps: 1, mixing lime with an ethanol solution, carrying out digestion reaction, and standing and settling lime slurry after the digestion reaction is finished; 2, carbon dioxide gas is introduced into the lime slurry for a carbonization reaction, vaterite-shaped calcium carbonate is generated, and different composite crystal growth inhibitors are added during carbonization to regulate and control the crystal form of calcium carbonate; and 3, after the carbonization reaction is finished, carrying out suction filtration, drying and crushing to finally obtain the monodisperse vaterite-shaped calcium carbonate with uniform particle size. The raw materials are low in obtaining cost, the preparation process is simple and easy to adjust, the calcium carbonate particles are regular in morphology and uniform in particle size distribution and are monodisperse products, the method for preparing the calcium carbonate with the special morphology is widened, and the prepared vaterite-shaped calcium carbonate is expected to be applied to the special fields such as the food industry and the pharmaceutical industry.
Owner:GUANGXI HUANA NEW MATERIALS TECH CO LTD

Partitioned coating method for space reflector

The invention relates to the technical field of optical coating, in particular to a partitioned coating method for a spatial extreme ultraviolet band reflector, which aims to coat different or same films on more than two or more regions at certain intervals. The method comprises the following steps: firstly, designing more than two or more EUV wave band multilayer films, wherein the structures and the central wavelengths of the EUV multilayer films can be different; then designing an adjustable shielding coating tool which is in direct contact with the mirror surface of the reflecting mirror; the EUV wave band is plated on the surface of the cleaned reflecting mirror in a regional mode through an adjustable shielding film plating tool; and finally, testing the reflecting mirror after film coating. According to the invention, the boundary of the coating area is clear and has no shadow effect, and stray light crosstalk is reduced; and the film is good in film forming quality, small in film stress, excellent in optical performance, good in film firmness and high in space environment adaptability.
Owner:CHANGCHUN INST OF OPTICS FINE MECHANICS & PHYSICS CHINESE ACAD OF SCI

Adaptive optimization method and system for parameters of digital analog optical fiber wireless forward transmitting terminal based on reinforcement learning

The invention relates to the technical field of optical fiber wireless communication and mobile forward transmission, in particular to a digital analog optical fiber wireless forward transmission transmitting end parameter self-adaptive optimization method and system based on reinforcement learning. Decomposing the normalized wireless signal into digital parts of all levels and a final analog residual error part through multi-level modulation, and transmitting the digital parts and the final analog residual error part in a time division multiplexing manner; a receiving end represents inter-symbol interference caused by link nonlinear distortion and bandwidth limitation, and the signal-to-noise ratio of a recovery signal is calculated based on the error vector amplitude. Parameters of a transmitting end are optimized and modeled into a Markov decision process, states comprise rounding factors, scaling factors, constellation forming factors and pre-equalization filter tap coefficients of all levels, and actions are discrete increments of the parameters; a reinforcement learning agent is constructed by using a dual deep action value network, and training is updated in combination with experience playback and a target network. The reward function comprises a threshold satisfaction item and a signal-to-noise ratio change item; parameters are interactively updated in the training stage, and online self-adaptive adjustment is performed in the reasoning stage.
Owner:FUDAN UNIVERSITY

Photoresist composition and preparation process thereof

The invention provides a photoresist composition and a preparation process thereof. The photoresist composition is prepared from the following raw materials in parts by weight: 80.0 to 100.0 parts of solvent, 20.0 to 40.0 parts of resin, 3.0 to 5.0 parts of photoinitiator, 1.0 to 3.0 parts of light stabilizer, 1.5 to 2.5 parts of tackifier, 1.0 to 1.5 parts of monomer, 1.0 to 2.0 parts of acid diffusion inhibitor and 0.5 to 2.0 parts of flatting agent. Wherein the acid diffusion inhibitor is prepared from perfluorocyclohexane, benzo-15-crown ether-5 and L-ethyl lactate, and the mass volume ratio g / mL of the perfluorocyclohexane to the benzo-15-crown ether-5 to the L-ethyl lactate is (2-3): (1-2): (1-2). According to the acid diffusion inhibitor disclosed by the invention, through the combined action of different mechanisms of the perfluorocyclohexane, the benzo-15-crown ether-5 and the L-ethyl lactate, the diffusion speed of acid is slowed down, so that the boundary between an exposure region and a non-exposure region is clearer, and the resolution ratio of photoresist is improved. By modifying the methacrylic resin, the surface energy and polarity of the methacrylic resin are changed, and the solubility of the resin after exposure is remarkably changed, so that the photoresist is easier to remove in a developing solution, and the definition and resolution of a pattern are improved.
Owner:GUANGDONG KEYOU MATERIAL TECH CO LTD

Pole piece production equipment and pole piece production process method

The invention relates to the technical field of battery production, and discloses pole piece production equipment and a pole piece production process method, and the pole piece production equipment is suitable for processing a continuously moving current collector; the pole piece production equipment comprises a first coating head and a second coating head, wherein the first coating head is suitable for coating a current collector with first slurry; the first drying oven is arranged at the downstream of the first coating head and is suitable for drying the current collector coated with the first slurry; the second coating head is arranged at the downstream of the first drying oven and is suitable for coating the current collector with second slurry, and the second slurry is located on at least one side of the first slurry in the width direction of the current collector; and the second drying oven is arranged at the downstream of the second coating head and is suitable for drying the current collector coated with the first slurry and the second slurry. According to the pole piece production equipment provided by the invention, the conditions of permeation and material mixing on the edges of the ceramic slurry and the positive active material slurry when the ceramic slurry and the positive active material slurry are in a fluid state before drying can be avoided.
Owner:ZHAOQING XIAOPENG INTELLIGENT MFG RES INST CO LTD

Boundary guided attention based encoder-decoder medical image segmentation system and method

The application relates to a boundary attention guiding-based encoder-decoder medical image segmentation system and method, and belongs to the technical field of medical image processing.The application comprises an encoder, a decoder, a feature enhancement module and a boundary attention guiding module.Different scale features are obtained in the encoder by inputting a medical image, and the features are output to the decoder; the decoder completes medical image segmentation according to the features; the feature enhancement module is further connected after the output of the different scale features of the encoder, and is used for enhancing the semantic features of the encoder, combining semantic information of different receptive fields, and inputting the feature output of the encoder into the decoder again after the feature output of the encoder passes through the feature enhancement module; the boundary attention guiding module is arranged between adjacent decoder modules in the decoder, and is used for strengthening boundary information of the medical image.The application is excellent in generalization ability and robustness, and the obtained medical image segmentation result is clear in boundary and obvious in small target.
Owner:CHONGQING UNIV OF POSTS & TELECOMM

A hyperspectral remote sensing image classification method based on hybrid convolutional neural network

The present invention discloses a method for classifying hyperspectral remote sensing images using a hybrid convolutional neural network. The present invention comprises: 1. connecting 3D and 2D convolutional neural networks in series, adding parallel convolution blocks and non-local attention modules, and constructing a custom neural network; 2. dividing annotated hyperspectral images into a number of overlapping small data cubes as inputs of the model; 3. dividing samples into training sets and test sets according to a certain ratio; 4. inputting training set samples into the neural network in batches for training until the model is stable; 5. using the trained model to classify the hyperspectral image, evaluating the classification effect of the model through the pre-divided test set, and finally obtaining a classification result graph. The present invention makes full use of the rich spectral information and spatial information of the hyperspectral remote sensing image. At the same time, 1×1 convolution is used to compress the channel dimension of the feature map during the attention operation, thereby reducing the amount of calculation of the model and improving the training speed of the model.
Owner:HANGZHOU NORMAL UNIVERSITY

A data processing method and system for government affairs service

This invention discloses a data processing method and system for government services, relating to the field of data transmission technology. The method includes: obtaining jitter-related USB events from USB events of a government self-service terminal; obtaining jitter-related government business events from government business events of the government self-service terminal based on the jitter-related USB events; obtaining a disturbance chain based on the jitter-related USB events and the jitter-related government business events, and determining the identifier of the disturbance chain; defining a target disturbance chain based on the identifier of the disturbance chain, and obtaining a disturbance chain pointing label; classifying all government data units according to the disturbance chain pointing label to obtain disturbance-related government data units and secure government data units. This invention improves the reliability and controllability of the government data processing process.
Owner:HAINAN LIAOSHI TECHNOLOGY CO LTD