32results about How to "Guaranteed learning effect" patented technology

The invention discloses a tunnel fire thermodynamic response reasoning method based on MGraphFomer. The method comprises the following steps: constructing a tunnel structure fire thermodynamic response database; according to tunnel classification design and structural damage judgment principles, structural thermal response grade judgment standards of different types of tunnels are constructed in a self-adaptive manner by means of a branch convolutional neural network, and structural thermal response grades of different types of tunnels are combined according to the structural thermal response grade judgment standards. Obtaining a structural unit mechanical response coarseness judgment model; based on a heterogeneous graph processing method of double encoders and a structural unit mechanical response coarse-grained judgment model, a structural unit multi-scale response fine-grained model is constructed, and a tunnel fire structure thermal response digital-intelligent prediction model is obtained; and the tunnel fire structure thermal response is predicted by using the tunnel fire structure thermal response digital intelligent prediction model, and reasoning of the tunnel fire structure thermal response is realized. The thermal response prediction method is high in accuracy and higher in generalization ability.

The invention relates to a lightweight electrical equipment multi-target detection method and system, and the method comprises the steps: collecting multi-angle images of different types of electrical equipment, obtaining an experiment data set, constructing a Ghostnet lightweight network structure, introducing a CBAM attention mechanism and an SPPFCSPC convolution structure to improve a YOLOv8 model, and obtaining a multi-target detection model based on the improved YOLOv8 model. And training, verifying and evaluating the multi-target detection model based on the improved YOLOv8 model to obtain a multi-target detection optimal model for electrical equipment multi-target detection. According to the method and the system, the model complexity and the calculation amount can be reduced while the detection precision is ensured, and an effective solution is provided for multi-target identification detection of lightweight electrical equipment.

The invention discloses a tunnel structure response prediction and protection optimization method based on a mechanical response model, and the method comprises the steps: carrying out the training and optimization of an explosion structure response neural network through an explosion structure mechanical response prediction database, and obtaining a tunnel explosion structure mechanical response model; based on a tunnel explosion structure mechanical response model, tunnel structure responses under different tunnel anti-explosion geometric protection means are obtained, and a geometric protection damage data set is constructed by calculating tunnel physical damage and functional damage; training and optimizing the geometric protection damage neural network by using the geometric protection damage data set to obtain a geometric protection damage model; and based on the geometric protection damage model, geometric protection means optimization is realized according to a damage minimization principle. The method is high in efficiency, strong in generalization ability and high in precision.

The invention provides a dual-channel federal map learning method, device and system with privacy protection, and belongs to the technical field of federal learning. The method comprises the following steps: a client performs feature aggregation on a locally held sub-graph, acquires aggregated node feature information, and sends the aggregated node feature information to a server; the server randomly performs cross-client feature aggregation between the two client sub-graphs by using the aggregated node feature information; the client interacts with the server according to the cross-client sub-graph information, and global meta-model parameters are obtained in a meta-learning mode; the client updates local model parameters according to the parameters of the global meta-model; and the client updates local model parameters according to the support set of the local sub-graph nodes and uploads the local model parameters to the server, and the server performs fine adjustment on global meta-model parameters by adopting a federated average algorithm to obtain a global model. According to the invention, while the learning effect is ensured, the privacy protection capability of the data is greatly enhanced, and the overall efficiency and effect of federal learning are improved.

The invention relates to a voice emergency help seeking method for old people based on the Internet of Things, and belongs to the technical field of data processing specially suitable for administrative, commercial, financial, management, supervision or prediction purposes. Performing uniform segmentation on a time axis on an emergency help-seeking voice segment of a set elderly patient to obtain signal amplitudes of all parts; and using a deep neural network model to intelligently identify an emergency help-seeking type code value corresponding to the emergency help-seeking voice segment based on the Internet of Things data and each signal amplitude. The invention further relates to an old man voice emergency help-seeking system based on the Internet of Things. According to the method and the device, the technical problem that in the prior art, emergency help seeking type identification is difficult to carry out on fuzzy voice sent by a weak old person patient is solved, the artificial intelligence model can be introduced to carry out intelligent analysis on the fuzzy voice signal of the old person patient by using the Internet of Things data, and therefore the technical problem is solved.

The present application discloses an intelligent training system based on an end-to-end multimodal teaching large model, which relates to the field of training information technology, and solves the technical problem that the existing artificial intelligence training is difficult to answer the questions raised by the trainees in time, resulting in poor effect of intelligent training; it includes an interactive module: extracting training documents, personalized modification information, question information and test answers in the training materials; a training material generation module: generating training materials according to the training documents and personalized modification information; a question and answer module: generating answer information according to the question information; an evaluation module: generating an evaluation test paper according to the training materials and question information; generating a training evaluation according to the evaluation test paper; generating personalized training materials according to the training evaluation of the evaluation test paper; a training module: conducting automatic voice explanation training according to the training materials; answering according to the answer information generated by the question information; and generating personalized training materials according to the training evaluation of the evaluation test paper; and improving the training effect of intelligent training.

The invention relates to the technical field of adaptive interaction, and discloses a platform interaction interface adaptive adjustment method and system based on image recognition, and the method comprises the steps: obtaining a user form image; performing facial expression recognition according to the user form image to obtain a facial emotion component; performing body language recognition according to the user form image to obtain a body emotion component; performing calculation according to the facial emotion component and the limb emotion component to obtain a user emotion vector; adjusting the page layout of the platform according to the user emotion vector and a preset page layout automatic adjustment strategy; and calculating an attention score according to the user emotion vector, and adjusting the teaching content difficulty of the platform according to the attention score and a preset teaching content difficulty automatic adjustment strategy. The method has the following effects that the platform interaction interface can be adaptively adjusted according to the use emotion of the user, and the adjustment efficiency is improved.

The invention discloses a lightweight multi-stage point cloud classification method based on a fusion Transform network. The lightweight multi-stage point cloud classification method comprises the steps of original point embedding module construction, local feature extraction, global feature aggregation and multi-stage learning. The network is composed of an efficient feature aggregation branch and a Transform branch, the efficient feature aggregation branch is provided with a learnable position code, and position information which is often neglected by most point cloud classification networks can be fully captured. Sampling points with similar characteristics are clustered by the Transform branch, so that the balance between capturing long-range dependence and calculation complexity is realized. Meanwhile, an extrusion-excitation module is introduced into original point embedding, so that representation of point cloud features is enhanced, and subsequent point cloud learning performance is improved. And a four-stage learning strategy is adopted, so that the understanding capability of the network on the point cloud is effectively enhanced. A large number of experiments prove that excellent classification performance can be achieved while the number of parameters is kept low.

This invention relates to the fields of machine learning and edge computing technology, specifically to a blockchain-based trusted distributed hybrid expert model edge computing system. The system includes a task publisher for publishing learning tasks, which may be training or inference tasks; a blockchain network for receiving learning tasks and evaluating expert weights to obtain sparse activation expert index results; a blockchain network for receiving learning task processing results from the edge layer and achieving consensus; a gating network deployed within the blockchain network; an edge layer for downloading the corresponding expert model from the storage layer based on the sparse activation expert index results to process the learning tasks and uploading the processing results to the blockchain network; and a storage layer for storing all expert networks, each with a CID. This invention significantly improves the defense capability of the distributed MoE network against malicious edge node data tampering attacks, thereby ensuring the learning performance of the MoE system.

The invention relates to a management method and system based on master data, and relates to the field of electric digital data processing. The method comprises the following steps: adopting an AI analysis model to intelligently analyze each access frequency of a target person in each operation system of a current enterprise in a current time segment on the basis of multiple person associated information of the target person and flow associated data of the target person in multiple historical time segments before the current time segment; and determining the importance level of the target person in the current time segment based on the cumulative value of each time. The invention further relates to a management system based on the master data. According to the invention, in order to solve the technical problem that it is difficult to configure the main data management strategy of each enterprise personnel in the future time segment in advance, the AI analysis model is adopted to intelligently predict the active degree of each enterprise personnel in the future time segment, and then the importance level of the enterprise personnel in the future time segment is identified. Therefore, the technical problem is solved.

The invention relates to a toll station lane dynamic configuration method and system based on data collaboration, and belongs to the technical field of traffic control. A cloud computing server is adopted to intelligently predict the total number of departure vehicles of different charging types in a current time segment of a target toll station according to an intelligent traffic prediction model established by an edge computing network element based on basic data collected by an Internet of Things sensor, and then the number of channels of different charging types is configured in advance; the invention further relates to a toll station lane dynamic configuration system based on data collaboration. According to the invention, aiming at the technical problem of poor traffic control effect caused by lack of basic data when an electronic toll channel and a manual toll channel of a toll station are dynamically configured at any time in the prior art, the number of toll channels of different types can be configured in advance according to the intelligently predicted total number of departure vehicles of different toll types; therefore, the technical problem is solved.

The present invention relates to an elderly voice emergency rescue method based on the Internet of Things, belonging to the field of data processing technology specifically suitable for administrative, commercial, financial, management, supervision or prediction purposes, and the method comprises: when receiving a fuzzy signal of a voice for help, evenly dividing the emergency rescue voice segment of a set elderly patient on the time axis to obtain the amplitude of each signal; using a deep neural network model to intelligently identify the emergency rescue type coding value corresponding to the emergency rescue voice segment based on the Internet of Things data and the amplitude of each signal. The present invention also relates to an elderly voice emergency rescue system based on the Internet of Things; through the present invention, in view of the technical problem in the prior art that it is difficult to identify the emergency rescue type of the fuzzy voice emitted by the weak elderly patient, an artificial intelligence model can be introduced to use the Internet of Things data to perform intelligent analysis on the fuzzy elderly patient voice signal, thereby solving the above technical problem.

This invention relates to a method for simulating and analyzing rainfall-induced landslide geological disaster scenarios, belonging to the field of image synthesis, and more specifically to the field of 3D geographic modeling for computer mapping. The method includes: collecting customized model data of a given mountain at the current moment based on a 3D graphical model of the mountain; and using a landslide intelligent prediction model to intelligently predict whether a landslide geological disaster will occur on each side of the given mountain within the next time interval, starting from the current moment, based on the customized model data. This invention also relates to a rainfall-induced landslide geological disaster scenario simulation and analysis system. Through this invention, the technical problem of difficulty in predicting future landslide data for landslide-prone physical entities such as mountains is addressed by intelligently predicting whether a landslide will occur on each side of the given mountain within a future time interval based on a 3D graphical model of the mountain at the current moment, thereby solving the aforementioned technical problem.

The invention relates to an intelligent distribution system for logistics order resources. The intelligent distribution system comprises an information acquisition part, a data capture part, a multi-layer building mechanism, a prediction processing mechanism, a store configuration mechanism, a prediction processing mechanism, a store configuration mechanism and the like. The logistics management level can be improved.

The invention relates to an image enhancement method and system based on an image processing chip, and relates to the field of general image data processing or generation. The method comprises the steps that an AI prediction model is adopted to intelligently predict the signal-to-noise ratio of an enhanced image obtained after an image to be enhanced is subjected to image enhancement processing of a target image enhancement algorithm based on various collection information of an image processing chip, multiple analysis parameters of the image to be enhanced and algorithm data corresponding to the target image enhancement algorithm; and selecting an optimal image enhancement algorithm of the to-be-enhanced image based on the signal-to-noise ratio of each part corresponding to each image enhancement algorithm. The invention also relates to an image enhancement system based on the image processing chip. According to the invention, aiming at the technical problem that respective optimal image enhancement algorithms of different images cannot be predicted, the AI prediction model with a customized structure can be adopted to predict the image enhancement effect obtained by the image processing chip by adopting any type of image enhancement mode, so that the technical problem is solved.

The application relates to a seamless switching and context synchronization system and method in a customer service man-machine cooperation process, and deep learning is provided in the application, and more particularly relates to the field of electric digital data processing. The system comprises: a big data acquisition device for acquiring big data track communication content of each past time segment of a set e-commerce application before a current time; and a data analysis device for intelligently analyzing the number of intelligent robots and the number of artificial customer services required for seamless switching in the current time segment based on the big data track communication content by using an intelligent customer service data analysis model. By the application, in the face of the technical problem that seamless switching in the customer service man-machine cooperation process is difficult to achieve in different time segments, a deep learning model can be used to intelligently analyze the number of various customer services required for seamless switching in each time segment according to various types of basic information acquired by big data, so that the above technical problem is solved.

The invention relates to a suspicious clue automatic monitoring data analysis method based on artificial intelligence, and belongs to the field of electrical digital data processing, and the method comprises the steps: executing a set number of each learning action on a convolutional neural network to obtain a suspicious clue automatic monitoring model; intelligently analyzing whether a target monitoring water surface belongs to a polluted water surface or not and a suspected distance from a pollution source to a current monitoring point by adopting a suspicious clue automatic monitoring model based on time-sharing dynamic water quality related data and time-sharing dynamic visual data; according to the invention, the technical problems that environmental pollution monitoring data analysis results are not rich enough and the analysis process is complex are solved; an automatic suspicious clue monitoring model designed in a targeted structure can be used for carrying out intelligent data analysis on suspicious clues such as whether environment pollution exists in a fixed water area and the distance from an environment pollution source to the fixed water area based on time-sharing dynamic water quality data and time-sharing dynamic visual data, so that the technical problems are solved.

The present invention relates to a method and system for dynamically configuring lanes at toll stations based on data collaboration, and belongs to the field of traffic control technology. The method comprises: in a network security data information transmission environment, a cloud computing server is used to intelligently predict the total number of outbound vehicles of different charging types in the current time segment of a target toll station based on the basic data collected by the Internet of Things sensor according to the intelligent traffic prediction model built by the edge computing network element, and then configure the number of channels of different charging types in advance; the present invention also relates to a system for dynamically configuring lanes at toll stations based on data collaboration. Through the present invention, the technical problem that the lack of basic data when the electronic toll channels and manual toll channels of toll stations are dynamically configured at any time in the prior art leads to poor traffic control effects, and the number of toll channels of different types can be configured in advance according to the total number of outbound vehicles of different charging types intelligently predicted, thereby solving the above technical problems.

The present application relates to the technical field of power consumption prediction, and particularly relates to a regional power consumption intelligent prediction method based on multi-source heterogeneous data, which comprises the following steps: determining the power consumption prediction influence tendency for the next preset monitoring period based on the power change characteristic value; when it is determined that the next preset monitoring period is a strong influence tendency for power consumption prediction, determining that the power consumption is corrected by the elastic influence weight coefficient, and determining whether the prediction for the power consumption is qualified based on the energy change characteristic value; when it is determined that the prediction for the power consumption is abnormal, correcting the prediction parameter for the power consumption in the next preset monitoring period based on the regional industry change quantitative value, including identifying the main analysis data set, determining whether to increase the model training set or separate the prediction model based on the period quantitative value of the main analysis data, and performing targeted analysis and prediction on the power consumption according to the specific situation of the power elasticity coefficient of the region to be predicted, thereby improving the prediction efficiency of the power consumption in the region.

The invention relates to a charging optimization method for a starting power supply system, and belongs to the field of power transmission, conversion, distribution and control equipment, and the method comprises the steps: operating an AI prediction model to obtain a success and failure identifier of a next starting vehicle battery of the starting power supply system; and when the successful failure identifier marks that the starting of the power supply system fails, forbidding the starting of the power supply system and executing an early warning operation corresponding to the next starting failure, otherwise, continuing to enable the starting of the power supply system. According to the invention, for the technical problem that the charging optimization space is limited because the future starting success or failure of the starting power supply system is difficult to predict, the success and failure of the next starting of different starting power supply systems are predicted by adopting the basic data screened in a targeted manner and utilizing the AI prediction model designed by a customized structure; and on the basis of prediction, the charging mode for starting the power supply system is optimized, so that the technical problem is solved.

The invention belongs to the technical field of educational resource management, and particularly relates to an educational resource management system and method based on the Internet and a storage medium, and the system comprises a platform end, a user end and a supervision end. The system integrates, classifies and stores various education resources through the resource management sharing module and updates the resources, the user management module manages user information of teachers, students, parents and administrators and carries out identity verification and authority control on all users, and the learning recommendation module can carry out targeted learning planning. The system helps students to learn and improve scores more efficiently, monitors the learning communication behaviors of the students through a behavior monitoring and analysis module, judges the learning performance of the corresponding students in a detection period, and sends a judgment result to the corresponding parents, thereby strengthening the subsequent learning supervision of the corresponding students in time, and improving the learning efficiency of the students. The subsequent learning efficiency and learning effect are ensured, and effective management of the education platform is realized.

The application provides a special management method and system for an education research platform, which comprises the following steps: checking the personnel identity of each logged-in person respectively, issuing corresponding research tasks and research permissions, obtaining the permission exercise data generated when each logged-in person executes the research task, constructing the research archives of each logged-in person in combination with the corresponding research permissions, constructing the research access path of the corresponding logged-in person in the research platform according to the permission exercise data, obtaining the corresponding research results of this time and transmitting them to the corresponding login page for display, analyzing the research reports of all the logged-in persons under different disciplines, and transmitting them to the platform homepage for display. The learning plan for each student in the education research platform is formulated, and the learning effect is analyzed, so that the learning effect of the student in the education research platform is effectively improved.

The invention provides a graph network information measurement method, a robust graph representation learning method and a graph data analysis method. The method mainly comprises the following steps: acquiring an adjacent matrix of a graph network topology; calculating an edge curvature of the graph network topology based on the adjacent matrix; and taking the edge curvature as graph network information measurement, and judging whether the related sub-topology has topology defects or not according to a network information measurement theory. According to the method, the edge curvature is used as network information measurement, and graph topology defect detection can be accurately realized. According to the robust graph representation learning method, the feature representations of the network topology after modification and the network topology before modification are extracted respectively, and the feature representations are integrated to obtain a representation learning result of the network topology, so that the stability and the accuracy of graph representation learning based on the graph neural network can be remarkably improved; the over-compression phenomenon of graph representation in the message passing process is reduced, and the representation learning performance is guaranteed.

The present invention discloses a method and system for generating traditional sachet patterns based on deep learning, the method comprising: obtaining a pattern sample set; preprocessing the pattern sample set to obtain an image set after contour detection; classifying the pattern sample set according to the image set after contour detection; then amplifying the pattern sample images of each category using affine transformation and making the number of images of each category consistent after amplification; unifying the image size of each category after amplification to obtain a pattern training sample set; inputting the pattern training sample set into a deep learning network for training to obtain a pattern training model. The present invention makes up for the lack of the number of existing pattern sample sets to a certain extent; and greatly alleviates the problem of insufficient novelty and low output of current pattern designs.

The invention discloses a music teaching recommendation method and system based on a knowledge graph, and relates to the technical field of teaching recommendations, the method recommends a target node to a student according to each node music teaching resource in the knowledge graph and the learning record of the learned student in each node, and then according to the historical learning record of the student, the target node is recommended to the student. According to the method and the device, a music score recommendation mode of a student is confirmed, a target teaching music score is recommended to the student, and a learning period of the target teaching music score is confirmed according to a historical learning condition of the target teaching music score and learning conditions of the student in different periods, so that the adaptability of the music score and the learning period to the student is improved, and the music score recommendation accuracy and reliability are improved; personalized and targeted recommendation of the music score of the student is realized, the learning interest of the student is effectively stimulated, the learning requirements of different students are met, and the learning effect of the student is guaranteed, so that the music teaching recommendation effect is improved, and the satisfaction degree of the student is improved.

The present invention relates to a method for equipment space planning based on spatial graphical standards, and relates to the field of data processing systems or methods for administrative, commercial, financial, management, supervisory or predictive purposes. The method comprises: performing each learning operation on a feedforward neural network to obtain a feedforward neural network model; based on the three-dimensional values ​​of each spatial graphic corresponding to each side of the target equipment in the three-dimensional space and the three-dimensional coordinate values ​​of each point evenly spaced in the three-dimensional space corresponding to the reference internal space of the current cabinet, using the feedforward neural network model to intelligently predict whether the target equipment can be successfully placed in the current cabinet; through the present invention, in view of the technical problems of insufficient speed, efficiency and accuracy of equipment space planning, the feedforward neural network model can be used to intelligently predict whether the target equipment can be successfully placed in the current cabinet after multiple attempts based on multiple spatial graphical standard data, thereby solving the above technical problems.

This disclosure relates to an interactive method, apparatus, medium, and electronic device for learning resources. The method includes: in response to receiving a selection operation for a learning resource, displaying in a sidebar of an interactive interface a first display control corresponding to each learning segment included in the selected learning resource and a second display control corresponding to each learning stage included in the current learning segment, wherein the second display control is located below the first display control corresponding to the current learning segment; in response to receiving a selection operation for the second display control corresponding to a learning stage in the current learning segment, controlling the sidebar to exit display and displaying the resource content corresponding to the selected learning stage in the interactive interface. Through the above technical solution, users can select the desired learning stage after selecting a learning resource, thereby obtaining the resource content corresponding to the selected learning stage, resulting in a good user experience.

This application discloses a training method, usage method, apparatus, device, and medium for an image classification model, belonging to the field of artificial intelligence. The image classification model includes a feature extraction network and a multiple instance learning model. The method includes: acquiring a sample image set, wherein each sample image in the sample image set includes at least two instances; training the feature extraction network using the sample images in the sample image set through self-supervised learning based on contrastive learning, obtaining a trained feature extraction network; and training the multiple instance learning model using the sample images in the sample image set through multiple instance learning based on a mutual attention mechanism, obtaining a trained multiple instance learning model. The above scheme can reduce the computational complexity of the image classification model. The embodiments of this application can be applied to various scenarios such as cloud technology, artificial intelligence, smart transportation, and assisted driving.

The invention provides a special education long and short term goal formulation and evaluation system and a use method, and the system comprises an information collection module which is used for collecting the personal information of a child, and an IEP formulation module which is used for formulating a long term goal and a short term goal of the learning of the child. The teaching course resource module is used for arranging children to learn courses and providing teaching tools used in the courses; and the intelligent teaching aid learning tool module is used for assisting an educator to carry out auxiliary teaching. Short-term learning and long-term learning targets can be adjusted in time along with teaching development, the learning potential of children is brought into full play, individual demands of the children are met, and continuity and effectiveness of education are ensured.

The invention relates to an equipment space planning method based on a spatial imaging standard, and relates to the field of data processing systems or methods for administrative, commercial, financial, management, supervision or prediction purposes, and the method comprises the steps: carrying out each learning operation on a feedforward neural network to obtain a feedforward neural network model; the method comprises the following steps of: obtaining a three-dimensional coordinate value of each point uniformly spaced in a three-dimensional space of a reference internal space corresponding to a current cabinet based on the three-dimensional value of each space pattern corresponding to each side surface of target equipment in the three-dimensional space and the three-dimensional coordinate value of each point uniformly spaced in the three-dimensional space; intelligently predicting whether the target equipment can be successfully put into the current cabinet or not by adopting a feedforward neural network model; according to the technical scheme of the invention, for the technical problems of insufficient equipment space planning speed, efficiency and precision, the feedforward neural network model can be adopted to intelligently predict whether the target equipment can be successfully put into the current cabinet after multiple attempts based on the multi-term spatial graphical standard data, so that the technical problems are solved.