1572results about How to "Reduce demand" patented technology

The invention relates to a real-time fault detection and automatic repair system for a power grid. The real-time fault detection and automatic repair system comprises a multi-level sensing module, a data processing module, an automatic repair module and a monitoring and logging module. The multi-level sensing module collects data such as voltage, current, frequency and temperature in operation of a power grid by deploying multiple types of sensors, and sends the data to the data processing module. And the data processing module analyzes the collected data through an embedded intelligent algorithm, identifies potential faults and generates fault diagnosis results. And the automatic repair module executes repair operations of isolating a fault area, reconstructing a power grid topological structure and recovering normal power supply according to a fault diagnosis result. The monitoring and log module records the operation state of the power grid and the fault diagnosis and repair process, and provides fault analysis and performance evaluation functions through a user interface. According to the system, real-time detection, accurate positioning and automatic repair of power grid faults and comprehensive monitoring of operation data are realized, and the safety and reliability of power grid operation are remarkably improved.

The invention belongs to the technical field of image classification neural networks, and discloses a lightweight image classification neural network architecture system based on knowledge distillation. The invention provides a lightweight image classification neural network architecture system based on knowledge distillation. The lightweight image classification neural network architecture system comprises a data set preprocessing module, a teacher model preprocessing module, a subnet model construction module, a personalized training module, a real loss calculation module, a distillation loss calculation module, a total loss calculation module and a student model construction module. The invention provides a lightweight neural network architecture system based on knowledge distillation, and designs a brand new training mode by using the superiority of knowledge distillation for teacher model weight extraction: the knowledge distillation of teacher and student models is combined with a deep compression method; the system for constructing the lightweight network can be used for training various complex image classification neural networks so as to improve the network performance.

The invention belongs to the technical field of data detection, and particularly relates to an efficient data anomaly automatic detection and alarm system. The detection and alarm system comprises a sensor and data acquisition device, a data processing and analysis module, an alarm module, a monitoring center and visual interface, a communication and network module, a storage and backup module, an intelligent decision support module and a user feedback and adjustment module. According to the system, data can be monitored and analyzed in real time, abnormity can be quickly found, an alarm is given, manual monitoring requirements are reduced, data misjudgment is avoided, data accuracy is ensured through data cleaning and processing, an intelligent decision support module provides optimization suggestions based on AI, and a user can adjust rules to adapt to requirements. The system has the functions of data processing, anomaly recognition and trend prediction, the monitoring center visually displays the state, the storage and backup module ensures long-term storage of data, the automatic mechanism improves the response speed and the operation benefits, and the adaptability is high.

The invention discloses an intelligent bidding and tendering information identification method, which relates to the technical field of engineering management, and comprises the following steps: standardizing and preprocessing a bidding and tendering information document format; constructing a natural language processing model special for the bidding and tendering field; training a multi-level information extraction model based on deep learning; carrying out cross-document information integration and correlation analysis; performing anomaly detection and verification based on rules; semantic understanding and sentiment analysis; automatic subject matter description and technical specification extraction; a real-time dynamic monitoring and information updating mechanism; the invention discloses a graph-based decision support and recommendation system. According to the intelligent bidding and tendering information identification method, efficient and intelligent identification of bidding and tendering information is realized by constructing a natural language processing model specially aiming at the field of bidding and tendering. The bidirectional coding capability of BERT and the sequence modeling capability of LTM are combined, and a hierarchical attention mechanism is introduced, so that the model can accurately identify various key information in bidding and tendering documents.

The invention relates to an intelligent semantic analysis knowledge graph system for multi-source heterogeneous data. The system comprises a data preprocessing module used for preprocessing to-be-processed multi-source heterogeneous data to obtain format-standardized text data; the semantic analysis module is used for performing natural language processing on the text data to obtain words, and then performing vocabulary layer, syntactic layer and semantic layer analysis on the text data and the words to obtain structured data; the structured data comprises a triple consisting of entities, relationships and events and a network diagram of the entities and the relationships; the model structure mapping module is used for mapping the structured data into a predefined structure of a knowledge graph to obtain structured graph data; and the knowledge graph module analyzes entities and relationships in the graph data, performs reasoning to obtain potential relationships in the graph data by utilizing a logic rule and a machine learning algorithm, realizes cross-data-source and cross-document deep semantic association, and improves the efficiency of large-scale data processing and knowledge management.

The invention discloses a terminal data automatic backup and recovery method and system, and relates to the technical field of data backup and recovery, and the method comprises the steps: dynamically adjusting the priority, and generating a personalized backup strategy; capturing data change in real time and executing incremental backup; and encrypting the backup data for version recovery. According to the terminal data automatic backup and recovery method provided by the invention, accurate analysis and priority division are performed on the terminal data through the classification model, so that the backup plan can be dynamically adjusted according to the importance of the data in the backup process. Data change is captured in real time, incremental backup is executed, and the real-time performance and efficiency of backup are effectively improved. An incremental backup algorithm is adopted, only a file change part is recorded, and a change area of a data block is quickly positioned by utilizing hash value comparison, so that the requirement of a storage space is reduced, and meanwhile, the transmission overhead of backup data is reduced. According to the method, better effects are achieved in the aspects of backup real-time performance, backup accuracy, backup pertinence and backup efficiency.

The invention discloses a glass cold bending test electric loading device and a test method.The device comprises a base, a fixing support, a pressure loading test piece, a lifting component, a loading integrated piece and a control unit, the fixing support is adjustably installed on the base, the pressure loading test piece is arranged at the top end of the fixing support, and the lifting component is arranged on the loading integrated piece; the pressure loading test piece is used for providing extrusion stress for a glass cold bending test and outputting stress data, the lifting component is installed on the fixed support, the loading integrated piece is arranged below the pressure loading test piece and installed on the lifting component through a connecting part, and the loading integrated piece is used for providing various glass cold bending loading modes. And the pressure loading test piece and the lifting component are electrically connected with the control unit. The pressure loading test piece connected with the control unit is introduced, and the lifting component and the loading integrated piece are matched, so that the loading synchronism, precision and test safety are improved, the data acquisition accuracy is ensured, the test efficiency is improved, and the device is suitable for cold bending test loading of various different forms.

The invention provides a dynamic multi-link intelligent management and scheduling system, and relates to the technical field of network communication. According to the system, real-time monitoring, dynamic management and transmission data scheduling of a network link can be realized, and the problems of low bandwidth utilization rate and insufficient transmission efficiency in the prior art are solved. The method comprises the following steps: firstly, acquiring and monitoring state information of each link in real time through an intelligent agent technology below a link layer; and secondly, the system performs link admission control according to real-time monitoring data by using a dynamic regulation rule method, and dynamically judges whether a new relay link needs to be established and provides an optimal relay node selection when the data transmission bandwidth between the two nodes is insufficient. And finally, multi-link transmission data scheduling is carried out by adopting a mode of combining deep reinforcement learning and a heuristic scheduling algorithm, and the bandwidth utilization rate is improved to the maximum extent. On the whole, an efficient and intelligent network link management scheduling scheme is provided through combination of methods of intelligent proxy, dynamic regulation rules, deep reinforcement learning and the like, and the utilization rate of network resources and the quality of data transmission are effectively improved.

The invention relates to a multi-mode intelligent question and answer reasoning method and device based on knowledge distillation, and belongs to the technical field of intelligent question and answer. Extracting multi-modal features of the user data; a pre-trained intention recognition model conducts reasoning on the multi-modal features, reply content is generated, or corresponding operation is executed; the training method of the pre-trained intention recognition model comprises the following steps: acquiring multi-modal data; preprocessing the multi-modal data to obtain a multi-modal fusion vector; performing fine tuning training on the base large model by using the multi-modal fusion vector to obtain a fine-tuned large model; inputting the multi-modal data into the fine-tuned large model for intention classification, judging whether the classified intention slot position is complete or not, and if the judgment result is that the classified intention slot position is incomplete, supplementing the intention to generate final intention information; and performing knowledge distillation on the fine-tuned large model to obtain the model. The real-time performance and accuracy of intelligent question answering are improved.

The invention provides an evolutionary neural architecture searching method and system based on similarity agent assistance, and the method comprises the steps: 1, initializing an architecture population, and selecting an architecture with the best performance as an initial reference architecture; 2, constructing a graph neural network variant as a feature extractor; 3, constructing an agent model, and training the agent model through a joint loss function; 4, the high-potential architecture is reserved according to the fitness value, authenticity performance evaluation is carried out on the high-potential architecture, and an evaluation result is added into the training set; combining the current population with the high-performance architecture predicted and screened by the proxy model, and updating the population through an environment selection strategy; and step 5, repeating the step 3 and the step 4 until the population performance converges, and finally outputting a global optimal architecture. According to the method, the high-performance neural network architecture can be quickly searched under limited computing resources, and an efficient and intelligent novel solution is provided for automatic design and optimization of the neural network.

The invention provides a recommendation strategy generation method based on big language model enhancement and related equipment. The method comprises the steps of obtaining a recommendation task; inputting the recommendation task into an intelligent agent, and obtaining a recommendation strategy output by the intelligent agent; wherein the intelligent agent is obtained by training based on a sample recommendation task and a large language model, and the large language model is used for predicting user behaviors based on the sample recommendation task. According to the method, the big language model is combined with the reinforcement learning agent, so that the problem that an accurate recommendation strategy is difficult to quickly provide by a traditional method is effectively solved, and the efficiency and the adaptability of a recommendation system are improved while the accuracy and the individuation degree of the recommendation strategy are improved.

The invention discloses an image classification method based on an extensible quantum convolutional neural network, and relates to the technical field of image classification, and the method comprises the steps: carrying out the preprocessing of to-be-classified image data, and then carrying out the slicing processing; converting the slice data into quantum state information through angle coding; a quantum system is constructed, the quantum system comprises a quantum feature extraction part and a quantum classification measurement part, the quantum feature extraction part comprises a quantum variational circuit model with a quantum convolutional layer, and the model can perform feature extraction on quantum state information by using alternating single quantum bit gates and double quantum bit gates; training a quantum variational circuit model by using a part of quantum state information; and inputting the other part of quantum state information into the trained quantum variational circuit model, and measuring the output quantum state of the quantum variational circuit model. According to the method, the problems that traditional deep learning is high in computing resource demand, limited in quantum computer scale and the like are solved, and a novel and practical method is provided for efficient image classification.

The invention relates to a multi-modal target identification and tracking prediction method based on a laser radar and a binocular camera, and belongs to the field of robot / unmanned system perception identification. Comprising the following steps: preprocessing a target two-dimensional image acquired by a binocular camera and target point cloud data acquired by a laser radar to obtain a preprocessed two-dimensional image and point cloud data; s2, performing two-dimensional target detection and tracking detection based on the preprocessed two-dimensional image to obtain a two-dimensional target detection and tracking result; s3, performing three-dimensional target detection based on the preprocessed point cloud data and the two-dimensional target detection and tracking result to obtain a three-dimensional detection result; and S4, based on the pixel coordinates in the two-dimensional target detection and tracking result and the pixel coordinates of the vertexes of the target point cloud boundary in the three-dimensional detection result, target fusion tracking prediction is carried out, and the target ID, the position, the category and the target point cloud boundary are obtained. And accurate identification and tracking of the target are realized.

The invention discloses a construction method for governing and repairing desertification degraded grassland in an arid area, and relates to the technical field of sand governing. The method comprises the following steps: evaluating a native vegetation type, vegetation coverage, a soil degradation and desertification degree, a soil type and a water resource condition; performing classified measures according to natural conditions, and determining different types of dune (flowing, fixed or semi-fixed) treatment measures; construction preparation; grid sand barriers are laid on the windward sides of the moving dunes and the semi-fixed dunes, and sand stabilization operation is conducted in a mechanical and manual combined mode; digging pits to plant indigenous shrubs subjected to special treatment, and using rooting water retention bags to improve the survival rate; the granular organic fertilizer and the seeds are mixed and stirred according to a certain proportion and then sown into soil at a time, the operation frequency is reduced, and precise fertilization is achieved. Through effective sand prevention and control measures, the sand wind damage is reduced, the living environment of farmers and herdsmen is protected, and the living quality is improved.

The invention provides a liquid cooling module and a battery pack, the liquid cooling module comprises a second liquid cooling plate and a third liquid cooling plate, the second liquid cooling plate is internally provided with a flow channel, the third liquid cooling plate and the second liquid cooling plate are arranged at an interval, and the third liquid cooling plate is provided with a first cooling part; according to the liquid cooling device, the first cooling part is arranged to be communicated with the flow channel so as to guide the cooling liquid into the first cooling part and / or guide the cooling liquid out of the first cooling part, so that the cooling liquid can flow between the second liquid cooling plate and the third liquid cooling plate, the flowing path of the cooling liquid is simplified, and the requirement of a bent or overlong pipeline is reduced; and the internal space utilization rate of the battery pack is improved.

The invention relates to a preparation method and application of an inner-rigid outer-soft core-shell structure silicon dioxide blocking agent for marine drilling fluid, and belongs to the field of deepwater oil and gas drilling. N-vinyl-2-pyrrolidone, dimethyl diallyl ammonium chloride, acrylamide and 2-acrylamide-2-methylpropanesulfonic acid are polymerized and grafted on the surface of nano silicon dioxide to form the silicon dioxide blocking agent with a core-shell structure and a rigid inside and a soft outside; by means of rigid plugging of a nano material and deformation filling flexible plugging of a high-molecular polymer, shale micro-nano pores are efficiently plugged with rigid inside and flexible outside, so that the stability of a well wall is maintained, meanwhile, secondary generation of hydrates is effectively inhibited, and the safety and efficiency of deepwater drilling are improved; the plugging agent provided by the invention reduces the demand on a large number of thermodynamic inhibitors, not only saves the cost, but also has the characteristic of environmental protection.

The invention relates to the technical field of state monitoring, in particular to an intelligent cervical vertebra state data monitoring system which comprises a data monitoring prediction module, a cervical vertebra risk analysis module, a cervical vertebra intervention making module and an intervention effect evaluation module. According to the invention, through real-time data analysis and continuous monitoring of the posture and pressure change of the cervical vertebra, high-sensitivity tracking of the health condition of the cervical vertebra is realized, the system adopts dynamic probability estimation, the health state can be updated in real time, the accuracy of abnormal recognition is improved, and the accuracy of abnormal recognition is improved by integrating data analysis of the posture and the pressure. The system can identify potential health risks in an initial stage, early warning of risk factors enables medical intervention to be carried out more timely and effectively, the demand and cost of long-term treatment are reduced, a targeted intervention plan is adjusted according to real-time data, it is ensured that each intervention measure accurately matches the current demand of a patient, and the patient experience is improved. The intervention process is optimized, and the cervical vertebra rehabilitation speed and effect are improved.

The invention provides a hydropower station floating object detection method based on global relation modeling, and aims to improve the detection precision and efficiency of hydropower station water surface floating objects, and the method comprises the steps: collecting hydropower station water surface scene data through a multi-view device; extracting shallow multi-scale features through a lightweight convolutional network; modeling a global relationship between the floaters through a multi-head self-attention mechanism; carrying out the classification and positioning of floating objects through a target detection head; network training is performed through a multi-task loss function to optimize the model; and reasoning by using the trained floating object detection network, inputting a to-be-detected water surface image into the network, outputting the category and the position of the floating object, and finally generating a detection result. According to the method, the global relation modeling method is introduced, the detection precision of the hydropower station water surface floating objects is effectively improved, detection errors caused by insufficient local features in a traditional method are reduced, and the method has low calculation complexity and is suitable for real-time detection in practical application.

The invention provides a large model segmentation and deployment method and device based on cloud edge end collaboration and a medium, and the method comprises the steps: giving a pre-trained large model as input, and evaluating the performance and resource consumption conditions of the large model during cloud edge end deployment; determining a segmentation point of the large model according to a performance evaluation result and a resource consumption condition; decomposing the large model into a plurality of sub-models according to the segmentation points; deploying each sub-model at a cloud end and an edge end; and establishing a cloud side end cooperative work mechanism, so that the sub-models cooperatively work at the cloud side and the edge side. The method can reduce the demand for computing resources and improve the response speed while ensuring the model performance, and is suitable for the demand in the field of intelligent manufacturing.

The invention discloses a mask generation and image decoupling diffusion three-dimensional medical image generation method, relates to the technical field of image processing and analysis, and solves the problems of poor generalization ability, low calculation efficiency and high dependence on training data during image conversion in the prior art. The method comprises the following steps: acquiring a mask image, and processing the mask image by using a multi-condition diffusion probability model to obtain a multi-label mask sequence; screening the three-dimensional medical image data according to a screening condition, determining an informed slice set, and inputting the multi-label mask sequence and the informed slice set into a volume diffusion generation model to obtain a three-dimensional composite image; random noise is added to the three-dimensional composite image, a diffusion semantic refinement generation model is utilized, the three-dimensional composite image added with the random noise is refined, and a refined three-dimensional image is obtained; according to the invention, data decoupling and model generation volume consistency are realized, and a three-dimensional medical image with higher quality can be generated.

The invention relates to a construction method of an unmanned aerial vehicle remote sensing image classification model. Performing structured representation of the remote sensing image based on hierarchical feature decomposition; the method comprises the following steps: dividing an unmanned aerial vehicle image into image blocks of different levels through a pyramid decomposition algorithm, performing hierarchical combination for features of each level, and forming a multi-scale structured feature vector through a feature weighting mechanism; capturing dynamic features fused with a spatial topological relation: after a graph structure is constructed, performing feature aggregation and spatial relation information propagation on each node by using a graph convolutional network GCN, so that the features of each node comprise own information and the influence of fusing surrounding ground features; for multi-tense image data, a dynamic graph convolutional network D-GCN is introduced, modeling is carried out on spatial relations at different moments, and changes of ground objects along with time and influences of the ground objects on classification are captured; self-supervised clustering learning of ground feature features: for ground features which are difficult to distinguish, a soft clustering method is introduced, so that the ground features belong to a plurality of categories at the same time;

The invention discloses a multi-layer surface paint treatment process for an enameled wire, and relates to the technical field of enameled wire production. A multi-layer surface paint treatment process for an enameled wire comprises the steps of unwinding and stretching, annealing and softening, paint curing, inspection and winding and the like, in the paint curing process, a wire can be painted, baked and cooled through a paint coating mechanism, a drying oven and a cooling box in a paint coating device, and the surface of the enameled wire is coated with paint. And then the surface paint is conveyed to the feeding side of the painting mechanism again through the guide mechanism, so that a plurality of painting and curing procedures can be repeatedly carried out, and the coating of multi-layer surface paint is realized. Only one painting device is needed in the multi-layer surface paint coating process of the whole enameled wire, so that the requirement of production equipment for producing the enameled wire with the multi-layer surface paint on a production site can be reduced.

The invention discloses a respiratory training quantitative evaluation method, which comprises the following steps of: acquiring physiological parameter characteristics related to heart rate, respiration and brain function of a human body through a human body physiological data acquisition platform, and performing dimensionality reduction, dimensionless processing and weighted summation on the characteristics to obtain a respiratory training quantitative evaluation result. The method comprises the following specific steps: firstly, collecting human heart rate, respiration and brain function related physiological features through a physiological parameter collection platform, then calculating feature classification weight of each dimension through random forest and cross validation and removing redundant feature dimensions, and then removing dimensions of features in a feature subset through a dimensionless method; and finally, performing weighted summation on the feature data through the feature weight to obtain a respiratory training quantitative evaluation result. Quantitative evaluation is carried out on respiratory training, so that people can be helped to formulate more scientific respiratory training strategies and methods.

The invention provides a method and system for optimizing a commodity title and computing equipment, and the method comprises the steps: obtaining commodity information which comprises an original title of a commodity and a commodity attribute; obtaining related flow words associated with the commodity by using an algorithm recommendation system; based on the commodity information, first candidate traffic words used for being added into the original title are selected from the related traffic words through a large language model; based on the commodity information and the first recommended traffic word, utilizing a large language model to generate a second candidate traffic word used for being added into the original title; based on the commodity information, the first candidate traffic word and the second candidate traffic word, determining excluded risk traffic words by using a large language model to obtain a third candidate traffic word; and selecting an optimized traffic word from the third candidate traffic words by utilizing a large language model based on the commodity information, and adding the optimized traffic word into the original title, thereby generating a new title. According to the technical scheme, merchants can be better helped to screen drainage keywords, commodity titles can be more conveniently optimized, and commodity competitiveness is improved.

The invention provides a bidirectional identity authentication and encryption communication method based on an HART-IP protocol, and the method comprises the following steps: S1, respectively deploying the IP addresses of a control device and a field device and corresponding public and private key data to the control device and the field device through a configuration tool; s2, the control equipment sends an authentication request command to the field equipment, the field equipment analyzes and verifies the authentication request command and returns an authentication request response to the control equipment, and the control equipment analyzes and verifies the authentication request response and sends an authentication feedback command to the field equipment again; s3, after encrypting the session key, the control equipment sends a session key distribution command to the field equipment, and the field equipment analyzes, verifies and decrypts the session key distribution command to obtain the session key; and S4, the control equipment and the field equipment perform encryption and decryption operation on the sent and received message data based on the session key to realize encrypted communication between the control equipment and the field equipment.

The invention discloses a cultivated land interplanting forest identification method combining mixed pixel decomposition and deep learning. The method comprises the following steps: generating basic classification data by processing remote sensing image data; on the basis of historical monitoring data and classification standards, cultivated land interplanting forest sample labels are established, and a standardized sample set is constructed; calculating end member abundance and a cross end member contribution ratio by using a mixed pixel decomposition technology, and completing preliminary classification in combination with a region area; designing a deep learning network architecture, and performing model training and optimization by using a multi-scale and direction feature extraction technology; automatically extracting spectrum, texture and spatial features through the model, and fusing to generate a classification result; and performing performance evaluation on the model by adopting a verification set and a test set, and optimizing classification parameters. According to the invention, accurate classification and identification of cultivated land interplanting forest ground feature targets are realized.

The invention provides an image enhancement system and method and a medium. The system comprises (N-1) row cache FIFO (First In First Out) which are connected in series, N row cache registers, a Demosaic module, a G component cache FIFO and a G component cache register. According to the invention, by reasonably controlling the read-write of the row cache FIFO and the register, at least five pixel points are output all the time under the condition that any number of pixel data is transmitted for each row of image data of each clock, so that the calculation requirement of the Demosaic algorithm is met. Under the condition of balancing and optimizing the combinational logic series to be executed in each clock period and the occupation of the total register, in the pipeline design of the Demosaic algorithm, the finiteness of hardware internal resources and the time sequence convergence problem of high clock frequency are fully considered, and by combining the combinational logic series optimization and the pipeline series optimization, the time sequence convergence problem of the high clock frequency is fully solved. In this way, better occupation of various resources of the FPGA is achieved, and efficient calculation is achieved. Based on the calculation strategy provided by the invention, the image enhancement process is finally realized, and the high-quality color display image is obtained.

The embodiment of the invention provides a dynamic access control method and device based on a security gateway and trust evaluation and a related product. The dynamic access control method based on the security gateway and trust evaluation comprises the following steps: closing an external access permission of an application and analyzing a domain name of the application to a security gateway address based on the security gateway, so that all access requests are accessed through the security gateway; whether an access control strategy is triggered or not is judged according to the risk assessment result and a pre-configured access control triggering rule, and if the access control strategy is triggered, the access control strategy is issued to the security gateway; and the security gateway receives the access control strategy in real time, intercepts an access request of a client to a target application, performs matching judgment on the content of an access request message according to a dynamic strategy, and executes a corresponding control action on the access request according to a strategy rule if a preset triggering condition is met so as to realize dynamic access control. And the access is safer and more flexible, and the self-adaption and the expandability are higher.

The invention discloses a composite material storage box forming mold and a manufacturing method thereof, and belongs to the technical field of composite material forming molds. An end socket connecting block, an end socket cylinder section connecting ring, a split type cylinder body assembly, a cap-shaped reinforcing rib and the like are manufactured in parallel and assembled to obtain the composite material storage box forming mold. The self-weight of the mold is greatly reduced while the outer molded surface of the combined mold is guaranteed, the mold filling and demolding processes are simplified, the end socket is left in the product after the product is formed, the strength and rigidity of the storage box are improved, and a stable connecting port can be provided.

The invention discloses a multi-modal large model automatic evaluation method and system based on Agentic Workflow, relates to the technical field of artificial intelligence model evaluation, and solves the technical problems that an existing large model evaluation method lacks image-text evaluation, is relatively high in subjectivity and is easy to cause illusion. The method comprises the following steps that S1, an input task is analyzed, and a structured task template is generated; s2, generating a multi-round multi-modal task flow chart based on the task template generated in the step S1; s3, based on the flow chart in the step S2, distributing the multi-modal large model component to each task node, executing a task and obtaining an intermediate result; s4, performing multi-dimensional scoring on the intermediate and final results in the step S3 by a plurality of scoring Agents, and fusing to generate an evaluation result; and S5, generating a feedback report based on the process data in the steps S1 to S4, identifying a model short plate, and providing an optimization suggestion. The method has the advantages of comprehensive evaluation of multiple modes, real-time feedback optimization and the like.