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

229results about How to "Easy to formulate" patented technology

Pollution research and water quality prediction method based on river segmented water environment capacity analysis

The invention discloses a pollution research and water quality prediction method based on river segmented water environment capacity analysis. The method comprises the following steps: accurately dividing a drainage basin into water environment capacity accounting units; obtaining a river water body image through satellite remote sensing; identifying and quantifying the positions and the intensities of point source and non-point source pollution sources in the drainage basin by utilizing the water quality parameters of the river obtained by the remote sensing image; establishing a river water quality prediction model; calculating the water environment capacity of each river reach under different surface water environment quality standards; comparing the estimated pollutant discharge amount with the water environment capacity under each water quality standard, and judging whether the pollutant discharge amount exceeds the capacity or not; and predicting future water quality change by using the model. According to the method, remote sensing, GIS and machine learning technologies are utilized, rapid analysis of the river segmented water environment capacity and accurate prediction of the water quality are achieved, dependence on a large amount of field sampling and complex software simulation is reduced, and the working efficiency is greatly improved.
Owner:南京市市政设计研究院有限责任公司

Remote rehabilitation management method and system based on virtual reality

The invention belongs to the field of remote virtual rehabilitation training, and particularly relates to a remote rehabilitation management method and system based on virtual reality. The method comprises the steps that firstly, the current health condition, the expected rehabilitation target and the needed period of a rehabilitation patient are predicted through a comprehensive evaluation algorithm and machine learning, secondly, a personalized training plan is intelligently recommended from a virtual rehabilitation training strategy library according to an evaluation result, and in the training implementation process, a sensor continuously monitors physiological indexes and micro-expression response of the rehabilitation patient; the rehabilitation progress is analyzed and evaluated through real-time data and fed back to a training strategy library in time, the training intensity and period are dynamically adjusted, and it is ensured that training is closely matched with the actual health state of a rehabilitative person; by fusing electronic medical record processing, virtual reality technology, machine learning prediction and visual emotion recognition, precise and intelligent management of rehabilitation training is realized, and rehabilitation efficiency and experience are remarkably improved.
Owner:YANGZHOU YIHANG MEDICAL EQUIP CO LTD

Intelligent precise irrigation system based on low-altitude remote sensing data driving

The invention discloses an intelligent precise irrigation system based on low-altitude remote sensing data driving. The intelligent precise irrigation system comprises a low-altitude unmanned aerial vehicle, a ground sensor network module, a meteorological data integration module, a space-time self-evolution feature map network module, a space-time adaptive attention network module, a context multi-arm tiger machine algorithm module and an automatic irrigation control module. The low-altitude unmanned aerial vehicle is used for acquiring remote sensing data of crops; the ground sensor network module is used for collecting real-time environment information; the meteorological data integration module is used for multi-modal data integration; the space-time self-evolution characteristic graph network module is used for generating a moisture demand distribution graph; the space-time adaptive attention network is used for optimizing an irrigation strategy; the context dobby machine algorithm module is used for dynamically selecting an optimal irrigation strategy; and the automatic irrigation control module is used for executing the optimal irrigation strategy and implementing differential intelligent irrigation management. According to the invention, precise and intelligent irrigation is realized by using the space-time self-evolution feature map network and the context multi-arm tiger machine algorithm.
Owner:CHANGCHUN DACHENGLIAN HI TECH CO LTD

Tunnel surrounding rock multi-mode intelligent fine grading method and system

The invention belongs to the field of tunnel engineering, and particularly discloses a tunnel surrounding rock multi-mode intelligent fine grading method and system, and the method comprises the steps: 1, constructing a sample library which comprises parameters while drilling of a tunnel face, a digital image and a corresponding tunnel face surrounding rock grade; 2, constructing a while-drilling parameter refined grading characteristic system; 3, constructing a surrounding rock intelligent grading model based on multi-modal data fusion; 4, dividing the rock mass in front of the tunnel face into a plurality of three-dimensional space unit bodies according to the tunnel contour and blast hole distribution characteristics; and 5, based on the surrounding rock intelligent grading model, finely and intelligently judging the grade of the tunnel face surrounding rock. According to the method, the intelligent surrounding rock grading model based on the multi-modal data fusion theory is constructed, the refined intelligent judgment of the tunnel face surrounding rock grade is realized, the precision and robustness of the refined intelligent judgment of the surrounding rock grade of the model are improved, the surrounding rock grading level in the construction stage is improved, more comprehensive reference is provided for a subsequent construction scheme, and the construction efficiency is improved. And tunnel intelligent construction is guided.
Owner:SOUTHWEST JIAOTONG UNIV

Exercise prescription recommendation method and device, electronic equipment and storage medium

The invention relates to an exercise prescription recommendation method and device, electronic equipment and a storage medium. The exercise prescription recommendation method comprises the following steps: constructing multiple agents at least integrated with a user agent, a doctor agent, a nutritionist agent and an exercise health agent, and providing an interactive interface to realize information sharing, competition or negotiation and decision and scheme generation between every two agents; pre-training is carried out based on the professional domain knowledge maps corresponding to the different agents, so that a multi-agent exercise prescription recommendation model is obtained; and inputting the obtained various related conditions of the metabolic disease patient into the exercise prescription recommendation model for reasoning to obtain a target exercise prescription. According to the invention, accurate and personalized exercise prescriptions can be formulated for patients with metabolic diseases.
Owner:SUZHOU YINGDI XINKANG NETWORK INFORMATION TECH +1

Thrombolysis prediction method based on deep learning and multi-modal fusion

The invention discloses a thrombolysis prediction method based on deep learning and multi-modal fusion, and belongs to the technical field of thrombolysis prediction, and the method comprises the steps: S1, collecting an MRI image of a patient with ischemic cerebral apoplexy, and constructing a cerebral apoplexy automatic segmentation model; s2, carrying out image registration; s3, lesion segmentation; and S4, thrombolysis prediction. According to the cerebral apoplexy automatic segmentation model based on the gating parallel Mama network, the long-distance dependency relationship in the image can be effectively captured, ischemic cerebral apoplexy lesions can be rapidly and accurately segmented, and the lesion identification efficiency and accuracy are greatly improved; according to the algorithm provided by the invention, the linear time complexity is realized, the computing resource consumption is greatly reduced, and the bottleneck of the prior art is broken through; according to the cerebral apoplexy thrombolysis prediction model based on the multi-modal data, scientific automatic decision support is provided for judging whether the patient is suitable for thrombolysis treatment or not, and the quality and safety of medical decision are improved.
Owner:TIANJIN UNIV

Distributed photovoltaic active power distribution network energy saving and loss reduction optimization method and system

An energy-saving and loss-reducing optimization method for a distributed photovoltaic active power distribution network comprises the following steps: performing unified scheduling on distributed photovoltaic output through an active coordination strategy, generating a photovoltaic output predicted value, correcting an error of the photovoltaic output predicted value, and forming a photovoltaic output curve; analyzing a real-time energy flow state in the power distribution network according to a photovoltaic output curve, dynamically adjusting a charging and discharging strategy of photovoltaic output and energy storage in combination with a load demand and a voltage level of a node of the power distribution network, and optimizing an energy transmission path; according to the analysis result of the photovoltaic output curve and the energy storage strategy, a multi-objective optimization model is constructed, the current state and the future operation demand of the power distribution network are comprehensively balanced, and the weight between optimization objectives is dynamically adjusted; and a multi-objective optimization result is applied to actual power distribution network operation, a self-learning closed loop is formed through real-time monitoring and feedback, photovoltaic output curve correction and strategy optimization, and the distributed photovoltaic active power distribution network is optimized.
Owner:STATE GRID JIANGSU ELECTRIC POWER CO LTD NANTONG POWER SUPPLY BRANCH

Pipeline dislocation alarm device for underground pipe gallery management

The invention belongs to the technical field of pipeline maintenance, and particularly relates to a pipeline dislocation alarm device for underground pipe gallery management, which comprises a pipeline detection robot main body, and detection cameras and water conveying pipes which are axially mounted on the outer side of the pipeline detection robot main body at equal intervals, one end of the pipeline detection robot body is rotationally connected with a cylindrical protective cover through a damping bearing, a rotary high-pressure flushing mechanism is arranged, the rotary high-pressure flushing mechanism can rotate and spray water at the same time, sludge at the dislocation position of a pipeline can be flushed in an all-dimensional mode, and the pipeline detection robot is more convenient to use. The situation that sludge adheres to the inner wall of the pipeline and influences the dislocation strength detection accuracy of the dislocation strength detection mechanisms on the dislocation position of the pipeline can be avoided, the two sets of dislocation strength detection mechanisms are controlled to contract inwards through the expansion mechanism after detection is completed, the dislocation strength detection mechanisms can be stored and protected, and the dislocation strength detection mechanisms can be prevented from being damaged.
Owner:POWERCHINA SEPCO1 ELECTRIC POWER CONSTR CO LTD

Cross-border logistics demand prediction method and system

The invention relates to the technical field of logistics management, in particular to a cross-border logistics demand prediction method and system, which can provide accurate and reliable prediction results and decision support for cross-border logistics enterprises. The method comprises the following steps: acquiring a destination cargo demand influence parameter set and a delivery place cargo supply influence parameter set; respectively performing feature analysis on the destination cargo demand influence parameter set and the shipping place cargo supply influence parameter set to obtain a destination cargo demand feature index and a shipping place cargo supply feature index; comparing the destination cargo demand characteristic index with the shipping place cargo supply characteristic index, extracting a minimum characteristic index, and marking the minimum characteristic index as a logistics demand characteristic index; obtaining the real-time cross-border transportation cost and the historical same-period cargo transportation characteristic quantity corresponding to the logistics demand characteristic index; and inputting the logistics demand characteristic index and the real-time cross-border transportation cost into a pre-constructed cross-border logistics demand influence analysis model as input.
Owner:SHENZHEN SHENCHUANGJIAN SUPPLY CHAIN CO LTD

Circuit and chip for supporting core particle parallel physical link through PIPE interface

The invention discloses a circuit and chip for supporting core particle parallel physical links through a PIPE interface, and the circuit comprises a data processing module which is located between the PIPE interface and a parallel PHY group, and comprises a data packaging module and a data analysis module, the data packaging module is used for generating a receiving direction data signal on the PIPE interface from data from the PIPE interface, compressing and packaging the receiving direction data signal and transmitting the receiving direction data signal to the parallel PHY group; and the data analysis module is used for analyzing the data from the parallel PHY group, recovering and restoring the compressed data, and then transmitting the data to an upper layer through the PIPE interface so as to send the data to a core particle interconnection controller on the PIPE interface. The invention aims to solve the problem of how to drive and use a plurality of parallel physical interfaces to carry out data transmission through a PIPE interface of a core particle interconnection protocol interface compatible with a PIPE interface on a physical layer, and data transmission is carried out through the PIPE interface and the plurality of parallel physical interfaces.
Owner:NAT UNIV OF DEFENSE TECH

Urban road traffic state prediction method based on macro-medium-micro hybrid simulation model, storage medium and equipment

The invention discloses an urban road traffic state prediction method based on a macro-medium-micro hybrid simulation model, a storage medium and equipment, and the method comprises the steps: collecting historical electronic police data and traffic channelization data, inputting the historical electronic police data and traffic channelization data into a constructed macro-medium-micro integrated multi-level hybrid simulation model, and carrying out the simulation, obtaining a road traffic state in each time period and traffic flow data at each moment; constructing a traffic flow parameter prediction model based on an MVF-FA-LSTM algorithm, inputting the traffic flow data into the traffic flow parameter prediction model, and predicting the traffic flow and the passing speed of the road; and constructing a traffic state prediction model based on a fuzzy C-means clustering algorithm, inputting the predicted traffic flow and passing speed of the road into the traffic state prediction model, and predicting the traffic state of the road. According to the method, unification of microscopic, mesoscopic and macroscopic simulation models is realized, and road flow and passing speed prediction is realized, so that road traffic states are divided, and a traffic management department is guided to command and make decisions.
Owner:SUZHOU IND PARK SURVEYING MAPPING & GEOINFORMATION CO LTD

Design cost optimization method, system and equipment of real estate project and medium

The invention discloses a design cost optimization method, system and device for a real estate project and a medium. The method comprises the steps that a plurality of product modular units are generated in a target engineering project; associating the product modular unit with a project unit identifier; in the target engineering project, obtaining a corresponding project product list library and a contract planning list; taking each product modular unit as a basic unit, determining a corresponding design cost influence factor according to the corresponding project product list library and the contract planning list, and performing engineering project cost budgeting to obtain a project cost budgeting result; in the planning and construction process, obtaining project construction data and actual construction cost of each product modular unit; obtaining a corresponding budget cost early warning interval based on the project unit identifier, judging whether the actual construction cost is within the corresponding budget cost early warning interval, and obtaining a budget cost evaluation result; according to the invention, the cost budgeting accuracy of the real estate project is improved.
Owner:GUANGZHOU XIANGSHI ENG CONSULTATION CO LTD

Power supply scheduling method and system based on distributed energy storage

The invention discloses a power supply scheduling method and system based on distributed energy storage in the technical field of power supply scheduling, and the method comprises the steps: obtaining an electric power system region range as a target scheduling region, collecting the region feature information in the target scheduling region, carrying out the energy storage division of an electric power system based on the region feature information, obtaining an energy storage data set, and storing the energy storage data set in a database; the energy storage data set comprises one piece of shared energy storage data and a plurality of pieces of energy storage data; dividing the target scheduling region based on the region feature information to obtain a plurality of target monitoring regions, each energy storage data corresponding to one target monitoring region; power demands in a power system area range are obtained, the power demands are divided based on the area feature information, a target energy consumption data set is obtained, and each target energy consumption data corresponds to a target monitoring area; transmission time can be saved, scheduling efficiency can be improved, a power scheduling plan can be better formulated, power resource configuration can be optimized, and operation efficiency of a power system can be improved.
Owner:GUOHUA ENERGY (JINGYANG) DIGITAL AGRICULTURE CO LTD

Multi-terminal series data transmission interaction system and method based on communication network

The invention belongs to the technical field of data transmission supervision, and particularly relates to a multi-terminal series data transmission interaction system and method based on a communication network, and the system comprises a data initiating terminal, a series data processing node module, a data receiving terminal, a communication monitoring management module, an emergency decision module and a background supervision terminal. Data to be transmitted are preliminarily packaged through a data initiating end and sent to a serial data processing node module, each data processing node in the serial data processing node module is responsible for a specific data processing task, and after processing of all the data processing nodes, a data receiving end receives and analyzes the data and sends the data to a server; the data processing and transmission efficiency is improved, the whole data transmission process of the data transmission link is monitored in real time through the communication monitoring management module so as to identify abnormity of the data transmission process, high efficiency, safety and reliability of data transmission are ensured, and the data transmission process is more controllable and predictable.
Owner:JILIN BUSINESS & TECH COLLEGE

Real-time monitoring and data processing method for reinforced concrete structure

The invention discloses a real-time monitoring and data processing method for a reinforced concrete structure, and particularly relates to the technical field of data processing, and the method comprises the steps: building a data anomaly detection influence model through obtaining the saline corrosion information, the dynamic load information of waves and inrush current and the attachment information of underwater organisms during the operation of an external sensor; measuring the influence of the dynamic load of saline corrosion, waves and inrush current and the attachment of underwater organisms on the existing data anomaly detection, comparing a data anomaly detection influence index with a preset data anomaly detection influence index threshold value, and when an anomaly detection influence signal is generated, constructing a data anomaly detection machine learning model, the data anomaly detection is optimized according to the influence degree of the data anomaly detection and the difference of the reinforced concrete structure data transmitted by the external sensor and the internal sensor, so that too frequent false report and missing report are avoided, the detection strategy is adjusted in time, and the accurate monitoring of the health state of the structure is improved.
Owner:ZHEJIANG QIANRUI TECH CO LTD

Unmanned aerial vehicle detection method based on ecological restoration technology

The invention discloses an unmanned aerial vehicle detection method based on an ecological restoration technology, belongs to the field of ecological restoration, and solves the problem that a monitoring camera cannot realize omnibearing coverage in the unmanned aerial vehicle monitoring process in the existing ecological restoration technology, and the method comprises the following steps: S1, before ecological restoration, carrying out large-scale investigation through an unmanned aerial vehicle; according to the method, partial core areas or areas worthy of attention are selected to carry out field detection, environmental elements (temperature, moisture and soil) are evaluated, a monitoring mechanism and a swing mechanism are arranged, the vegetation condition and terrain and landform data in the areas can be monitored in real time, a three-dimensional model of the areas is constructed through the monitored data subsequently, and the real-time monitoring of the areas is realized. According to the technical scheme, a corresponding ecological restoration plan can be conveniently made according to known data, the angle of the monitoring assembly can be flexibly adjusted in the monitoring process, the equipment can clearly and accurately recognize various conditions and data of the current area, the equipment monitoring accuracy is improved, and therefore the working quality of the equipment is improved.
Owner:INST OF MOUNTAIN HAZARDS & ENVIRONMENT CHINESE ACADEMY OF SCI

Power distribution network double-layer extension planning method considering carbon transaction and demand side management

The invention discloses a power distribution network double-layer expansion planning method considering carbon transaction and demand side management. The method comprises the following steps: acquiring data information of a power distribution network to be expanded; constructing a typical day scene based on a K-means clustering algorithm; the dynamic carbon emission intensity based on the carbon emission flow theory is adopted to calculate the carbon emission of the power distribution network; according to the time-of-use electricity price and the dynamic carbon emission intensity data of the power distribution network, the demand side electrical load is managed, and carbon transaction is carried out; establishing a power distribution network double-layer expansion planning model considering carbon transaction and demand side management; and upper and lower layer association of the power distribution network double-layer expansion planning model is converted into a single-layer model, linearization processing is carried out, and a solver is called for solving. According to the power distribution network double-layer extension planning method considering the carbon transaction and the demand side management, the carbon emission level of the power distribution network is effectively reduced, the system economy is improved, and a planning scheme which adapts to source load fluctuation and improves the operation stability of the power distribution network can be obtained.
Owner:CHINA THREE GORGES UNIV

MRI lumbar sagittal view image disease diagnosis method based on YOLOv8s algorithm

The invention belongs to the technical field of medical diagnosis, and discloses an MRI lumbar sagittal view disease image diagnosis method based on a YOLOv8s algorithm, and the method comprises the following steps: S1, screening collected MRI lumbar sagittal view image data, self-building a VOC2007 data set, and marking the data set; s2, performing data enhancement and other processing on the existing data set; s3, establishing a YOLOv8s deep learning neural network detection model; dividing the data set into a training set, a verification set and a test set according to a proportion of 8: 1: 1, training the YOLOv8s network until convergence, and obtaining a lumbar disease image detection model based on the YOLOv8s network and a weight file; and S4, testing the randomly extracted MRI lumbar vertebra sagittal view image of the patient by using the trained YOLOv8s detection model, outputting a detection result, and evaluating the detection result. According to the invention, the diagnosis of the MRI lumbar sagittal view disease can be realized, the detection precision and efficiency are greatly improved, and the diagnosis accuracy is improved.
Owner:QINGDAO UNIV OF SCI & TECH

Carbon sequestration capability value evaluation system and method of river and lake water ecosystem

The invention discloses a carbon sequestration capability value evaluation system and method for a river and lake water ecosystem, and the system comprises a data collection module which is used for collecting different types of vegetation samples in a target region and corresponding vegetation image data; the ground monitoring terminal comprises a data analysis processing module which is used for processing the collected vegetation samples to obtain carbon content data of different types of vegetation, obtaining vegetation biomass data of different types of vegetation according to the vegetation samples and the vegetation image data, and sending the vegetation biomass data to the ground monitoring terminal; calculating carbon sequestration capability data of different types of vegetation according to the vegetation biomass data and the carbon content data; the evaluation module evaluates the carbon sequestration capacity value of the target area based on the sum of the carbon sequestration capacity data of different types of vegetation, and the mode is beneficial to understanding the influence of the vegetation types on the carbon sequestration capacity value of the river and lake water ecosystem, facilitates the formulation of ecological protection policies, and improves the pertinence and effectiveness of the control and protection of the river and lake water ecosystem.
Owner:CHINESE RES ACAD OF ENVIRONMENTAL SCI

Data fragmentation transmission method and system based on logic strong isolation device

The invention discloses a data fragmentation transmission method and system based on a logic strong isolation device, and belongs to the technical field of information processing, and the method comprises the steps: S1, classifying to-be-transmitted data based on a first transmission demand to obtain a first classification result, and fragmenting the to-be-transmitted data based on the first classification result to obtain fragmented data; s2, establishing a data exchange channel between the data sending domain side and the data receiving domain side based on the logic strong isolation device; s3, encrypting the fragmented data based on the JDBC connection to obtain encrypted data, and storing the encrypted data in a database at a data receiving domain side based on a data exchange channel; and S4, obtaining recombined data from the database based on the identification features of the fragmented data, and obtaining data to be transmitted by a data receiving network domain side based on the recombined data. The problem that data stability and safety are poor due to the fact that data penetrating through a strong isolation device are not fully protected is solved.
Owner:ZHEJIANG HUAYUN INFORMATION TECH CO LTD

Document key information extraction method based on large language model

The invention discloses a document key information retrieval method based on a large language model, a retrieval result is output according to a multi-stage input document and a question input by a user, and the method comprises the following steps: constructing a domain knowledge graph; constructing and training a question extraction model, and obtaining key information and intention of the question input by the user through the question extraction model; retrieving in a knowledge graph according to the key information and the intention to form paragraph description of the key information; screening the document and splitting the document into a plurality of text blocks to obtain a plurality of text blocks most related to the content input by the user; obtaining a corresponding text block with the highest embedding similarity by using a large language model; and integrating and reasoning the corresponding text blocks and paragraph descriptions by adopting a large language model to obtain information about the user input question in the document. The method has the beneficial effects that the accuracy and efficiency of information extraction can be improved, specific requirements of different fields can be met, and powerful technical support is provided for document automatic processing and knowledge management.
Owner:STATE GRID JIANGSU ECONOMIC RES INST

Energy equipment data communication method based on block chain

ActiveCN119420784AEnsure transmission securitysafe storageSecuring communicationTarget controlEngineering
The invention provides an energy equipment data communication method based on a block chain, and the method comprises the steps: obtaining energy data of energy equipment which completes registration in a block chain network, and writing the energy data into a sub-chain of the block chain network based on the type of the energy data; the target energy equipment responds to a preset data interaction condition meeting the target control mode, initiates a first identity authentication request to the block chain network to perform first identity authentication, generates a data sharing request for the target type data based on the preset data interaction condition of the target control mode when the first identity authentication is successful, and performs second identity authentication; and when the access is successful, processing the data sharing request for the target type data by a sub-chain matched with the type of the target type data in the block chain network based on an intelligent contract, determining target shared energy data conforming to the access rule, and sharing the target shared energy data to the target energy equipment, therefore, energy data interaction between household energy devices can be realized safely and efficiently.
Owner:GUANGZHOU RIMSEA TECH CO LTD

Marketing scene process management system and method based on artificial intelligence

The invention discloses a marketing scene process management system and method based on artificial intelligence, and relates to the technical field of intelligent informatization management, and the system comprises a data collection module which is used for collecting data related to marketing activities, including consumer information and order information; the data storage module is used for storing the collected data and processing the data; the marketing analysis module is used for analyzing the collected data, portraying the user and analyzing the personal preference of the user; and the process optimization module is used for generating an optimization scheme of the marketing process according to the analysis result in combination with the original marketing process. According to the marketing scene process management system and method based on artificial intelligence, data related to marketing can be collected and analyzed, a worker can conveniently formulate a corresponding marketing process according to the interest of a user, and therefore more accurate and efficient marketing management is achieved.
Owner:SHANGHAI JINGSHUO NETWORK TECH CO LTD

System and method for characterizing radiation dose distribution in radiotherapy treatment plan

A method for a system for characterizing a radiation dose distribution in a radiotherapy treatment plan is disclosed. The radiotherapy treatment plan is used to treat an anatomical structure with a radiotherapy device. The system includes a processor and a user interface. The processor is configured to: receive data obtained from a radiation simulation at the anatomical structure under a nominal condition and one or more artificially applied uncertainty conditions; generating a plurality of dose distributions including dose-volume histograms (DVHs) using the received data, each DVH representing a tissue volume of the anatomical structure receiving a respective radiation dose, the DVHs including a nominal DVH corresponding to a nominal condition, and one or more uncertain DVHs corresponding to one or more artificially applied uncertain conditions; and determining a dose distribution characteristic based on statistical characteristics of the generated DVH. The user interface is configured to display the determined dose distribution characteristic to a user.
Owner:ELEKTA SHANGHAI TECH CO LTD

Clinical data classification method based on BERT, Bi-LSTM and attention mechanism

The invention relates to a clinical data classification method based on BERT, Bi-LSTM and an attention mechanism, and the method comprises the steps: collecting clinical data of a patient, and carrying out the tag processing; performing word segmentation processing by using a word segmentation device; bERT is used as an encoder, and feature representation of data after word segmentation is obtained; inputting the feature representation into a Bi-LSTM model to obtain an advanced representation of the data sequence; inputting the advanced representation of the sequence into an attention layer to calculate an attention weight; and multiplying the attention weight by the output of the Bi-LSTM, inputting a result into a full-connection layer to perform data classification, and performing data classification by means of a softmax activation function. According to the method, the BERT, the Bi-LSTM and the attention mechanism are combined, long-text clinical data of the patient can be better processed, and the long-distance dependency relationship in the text can be better understood, so that the clinical data of the patient can be better classified, the disease type of the patient can be better analyzed, and a doctor can be helped to accurately grasp the disease condition and make a treatment scheme.
Owner:SHENYANG INST OF COMPUTING TECH CO LTD THE CHINESE ACAD OF SCI

Method for measuring rigidity of paired main bearings of RV speed reducer

The invention discloses a method for measuring the rigidity of paired main bearings of an RV speed reducer. The method comprises the following steps of 1, customizing a bearing mounting tool according to the size of a bearing to be measured; step 2, inputting a test load spectrum and an axial measurement load on the measurement controller; 3, completing the assembly of the bearing installation tool and the two to-be-tested bearings, and connecting a tool upper cover with the top platform; 4, rotating the tool spacer ring and driving the outer rings of the two to-be-tested bearings to rotate, and eliminating gaps between the ferrules of the two to-be-tested bearings and the rolling balls; step 5, applying an axial load to the two to-be-measured bearings until the axial load is consistent with the axial measurement load set on the measurement controller, thereby completing the preloading of the two to-be-measured bearings; step 6, calculating the assembly height of the paired bearings; and step 7, taking the numerical value of the assembly height of the paired bearings as an original point, continuously applying an axial load to the two to-be-tested bearings, and finally drawing a stress-displacement curve by using test data. The method is high in measurement efficiency.
Owner:SHANGHAI UNITED BEARING

Movable tray structure for sugar painting machine

The invention discloses a movable tray structure for a sugar painting machine, and relates to the technical field of food processing, the movable tray structure is applied to supporting and conveying of sugar painting finished products in the sugar painting machine, the movable tray structure comprises a conveying device, a guide piece, a tray structure and an identification assembly, and the conveying device is installed on the inner bottom face of the sugar painting machine; the guide piece is mounted at the bottom of the sugar painting machine; the tray structure is connected with the conveying device, the conveying device drives the tray structure to slide along the guide piece, and two sets of detachable supporting pieces are arranged on the upper end face of the tray structure; the technical key points are as follows: the tray structure of the existing sugar painting machine is improved, so that the tray of the sugar painting machine has a bamboo stick recognition function and is used as a precondition for equipment operation, and the whole equipment is started after a bamboo stick is recognized, so that the bamboo stick can be effectively prevented from being adhered to the movable tray; and the steps of cleaning the sugar painting machine by workers can be effectively reduced, the using effect is good, and the good using prospect is achieved.
Owner:HEFEI ZANTONG INTELLIGENT TECH CO LTD

Subway passenger flow prediction method and equipment based on multi-scale space-time diagram attention network

The invention relates to a subway passenger flow prediction method and device based on a multi-scale space-time diagram attention network. The method comprises the steps of collecting passenger flow data of subway OD pairs, subway line network structure data and land use type data; constructing an OD-level urban rail transit knowledge graph based on the collected data, and calculating node features of the knowledge graph, including calculating historical passenger flow features, calculating land use information, calculating accessibility of OD pairs and calculating correlation of the OD pairs; inputting node features of the knowledge graph into a multi-scale space-time diagram attention network, performing diagram filtering on the node features, and updating the node features of the knowledge graph in combination with a result of a space-time attention mechanism; and inputting the updated node features into the multi-scale LSTM model, and outputting a multi-scale prediction result of the subway passenger flow. Compared with the prior art, the method has the advantages of being more accurate in prediction, beneficial to intelligent scheduling and the like.
Owner:SHANGHAI UNIV OF ENG SCI

Water quality evaluation WQI expert weight fusion method based on weighted extremum fusion algorithm

The invention discloses a water quality evaluation WQI expert weight fusion method based on a weighted extremum fusion algorithm, and relates to the technical field of water quality evaluation WQI. The method comprises the following steps: designing a single-factor evaluation algorithm; the importance degree of the water quality index is described by using three dimensions, and three different weighting methods are provided for the three dimensions; an algorithm, namely a weighted extremum fusion algorithm, which can fuse three weighting methods and realize subjective and objective combination is provided; by comprehensively considering the uncertainty degree of the water quality data, the index contribution degree and the regional key degree, the more scientific water quality evaluation method is provided, the method not only can accurately reflect the water quality condition, but also can provide a large number of adjustable parameters so that a user can carry out optimization according to specific requirements, and the user experience is improved. Therefore, water quality evaluation requirements of different regions can be better met, and the overall evaluation effect and practical value are improved.
Owner:周密