70results about How to "Achieve global optimization" patented technology

The invention relates to the technical field of power dispatching, and provides a multi-electricity energy storage participation power dispatching system based on an optimal charging demand, which comprises a dynamic price adjustment module, a multivariate data fusion module, a collaborative optimization intelligent dispatching module, an energy storage resource management optimization configuration module and a safety and reliability prompt module, a dynamic price adjustment module, a data fusion module, a collaborative optimization scheduling module, an energy storage resource optimization configuration module and a safety and reliability prompt module are integrated, global optimization of electric power resources is achieved through the intelligent and automatic technology, an intelligent bidding strategy is provided for dynamic price adjustment, a comprehensive information platform is constructed through data fusion, flexible scheduling is achieved through collaborative optimization, and safety and reliability are improved. The energy storage resource management platform monitors and optimizes configuration in real time, safety and reliability are prompted to ensure stable operation of the system, the system is intelligent, the automation integration degree is high, the power dispatching precision, the response speed and the economic benefits are remarkably improved, and the problem of insufficient economic benefits in the prior art is solved through the technical scheme.

The invention discloses a dynamic control method for a complex curved surface five-axis numerical control machining tool vector, which comprises the following steps of: scanning the surface of a target workpiece by using a high-precision three-dimensional scanner to obtain point cloud data; based on the point cloud data, establishing a mathematical model S (u, v) of the target complex curved surface; a mathematical model based on the target complex curved surface; establishing a comprehensive dynamic model of the five-axis numerical control system; optimizing the initial tool posture in combination with the kinetic model; collecting dynamic operation data of the cutter and the workpiece; on the basis of the dynamic operation data, a model predictive control (MPC) algorithm is used for adjusting the machining path and the cutter direction in real time; according to the method, a comprehensive dynamic model of a five-axis numerical control system is established, dynamic coupling effects of a machine tool kinematic chain, cutter flexibility, cutting force, a temperature field and a stress field are brought into a unified framework, and complex dynamic behaviors are accurately described and controlled through multi-body dynamic modeling and an optimization algorithm.

The invention relates to an image segmentation method for a burn wound area, and the method comprises the steps: collecting burn wound images under different spectrums through a multispectral imaging device, and enabling the burn wound images to comprise a plurality of wavebands, namely visible light, infrared light and ultraviolet light; for each spectral image, analyzing reflection and absorption characteristics of a burn area under different wavebands by using optical characteristics of biological tissues, and marking a potential burn area; designing a self-adaptive algorithm, and dynamically adjusting parameters, including exposure and light intensity, of imaging equipment based on feedback of different spectral characteristics; the feature contrast of the burn area is enhanced; preliminarily delimiting the boundary of the burn region according to the enhanced spectral image through a threshold segmentation or region growing algorithm; inputting the multispectral image into a hierarchical processing model, and decomposing different tissue levels of the burn area including epidermis and dermis based on different wave band or color information; on each layering result, a feature extraction algorithm including texture analysis and edge detection is used for extracting detail features of each layer.

The invention relates to the technical field of high-speed motor intelligent control, and discloses a high-speed motor intelligent control method and device for a cooling fan, and the method comprises the steps: carrying out the parameter collection of a high-speed fan motor, and obtaining a motor basic parameter group; performing motor overmodulation boundary calculation to obtain a linear PWM region threshold value and an overmodulation region threshold value; performing multi-objective optimization operation on the voltage distortion degree, the current harmonic quantity and the system loss to obtain an initial control parameter group; performing current harmonic analysis to obtain a current harmonic compensation value and a system delay compensation value; performing online state evaluation operation to obtain a PWM control parameter correction value, a current controller parameter correction value and a voltage vector distribution correction value; and performing fault feature analysis on the running state of the high-speed fan motor, and outputting a fault phase isolation instruction and a fault-tolerant control instruction, thereby improving the overall running effect of the high-speed fan motor.

The invention discloses a CPU junction temperature control method and system, and the method comprises the steps: obtaining first temperature data collected by an internal temperature sensor and n pieces of second temperature data collected by n ambient temperature sensors disposed around, constructing a temperature field grid topological graph, and carrying out the calculation to obtain standardized temperature field data; carrying out first-order and second-order derivative calculation of time dimension, constructing spatio-temporal characteristic sequence data, and calculating to obtain topological characteristic vector data of the temperature field; the attention weight is adopted, and fusion feature vector data are obtained through calculation; calculating to obtain system stability index data by adopting a gating circulation unit network; constructing a workload feature vector, and performing classification processing to obtain load type index data; calculating to obtain multi-level early warning threshold set data; and a prediction model is constructed and optimized to obtain control parameter set data, and a final control instruction sequence is generated. According to the invention, the temperature control precision is improved, and the reliability and the use efficiency of the power utilization information acquisition terminal are improved.

The invention relates to the field of industrial automation and intelligent logistics, and discloses a vision-based operation management system, which comprises a visual perception module, a traffic modeling module, a path optimization module, a scheduling optimization module and a feedback adjustment module, the visual perception module obtains image or video data in the operation area and extracts the dynamic state of the target equipment; the traffic modeling module constructs a dynamic weighted graph model based on the equipment state data; the path optimization module generates an optimal path of the equipment by solving a dynamic Hamiltonian path problem; the scheduling optimization module optimizes the passing sequence and speed of the equipment based on a game theory model; and the feedback adjustment module monitors the running state of the equipment in real time, detects abnormal conditions and dynamically adjusts modeling and optimization results. The method can effectively solve the problems of path conflict and traffic jam in multi-device collaborative operation, improves the operation efficiency and the resource utilization rate, has high real-time performance and adaptability, and is suitable for various scenes such as logistics storage, industrial production and intelligent transportation.

The invention relates to the technical field of traffic control, and discloses an expressway traffic prediction method and system based on a cloud side end, and the method comprises the steps: configuring data collection equipment of an expressway, and constructing a whole-region road sensing network of the expressway; configuring edge computing nodes of the expressway at the key hub, and obtaining a lightweight data packet of the expressway; according to the lightweight data packet, cloud storage data of the expressway is generated, and a traffic mode of the expressway is analyzed; identifying a traffic flow change trend and an abnormal traffic condition of the expressway, and setting an abnormity self-service processing mechanism and an intelligent auxiliary system of the expressway; defining a traffic management mode and a user interaction mechanism of the expressway, and generating a data sharing platform of the expressway; and in combination with the whole-region road sensing network, the traffic flow change trend and the data sharing platform, traffic prediction processing of the expressway is executed, and a prediction processing result is obtained. According to the invention, the prediction efficiency and accuracy of the expressway traffic prediction system can be improved.

The invention discloses a coal-based n-alkane phase change material optimization extraction method based on a data-driven model, and relates to the technical field of data processing. The method comprises the following steps: step 1, collecting experimental data of coal-based n-alkanes; constructing a coal sample feature vector based on the experimental data; 2, constructing a dissolution selectivity prediction model based on the coal sample feature vector and a dissolution parameter matrix of the coal sample; 3, constructing a multi-field coupling transmission equation; establishing a separation efficiency prediction model and a phase change performance prediction function according to a multi-field coupling transmission equation; establishing a crystallization process control equation, and solving the crystallization process control equation to obtain the crystallinity; step 4, constructing a multi-objective optimization function; solving a multi-objective optimization function to obtain an optimal process parameter combination; according to the method, the energy consumption and the production cost are remarkably reduced while the material purity and the extraction efficiency are ensured.

The invention discloses a street lamp group autonomous scheduling method and device based on an improved game theory and a medium, belongs to the technical field of intelligent street lamps, and is used for solving the technical problem that an existing street lamp scheduling method lacks comprehensive analysis capability of various environmental data and lacks dynamic adjustment and optimization capability. The method comprises the steps of collecting brightness data, traffic flow data and people flow density data of any street lamp node in an intelligent street lamp group through an Internet of Things technology, and constructing an environment data set; based on the environment data set, constructing an improved game theory model; based on the improved game theory model, determining an optimal operation strategy of any street lamp node; and sending the optimal operation strategy to a corresponding street lamp node, and indicating the street lamp node to adjust the illumination intensity and / or the on-off state based on the optimal operation strategy. According to the method, the energy efficiency and the response capability of the intelligent street lamp group are remarkably improved.

The invention discloses an intelligent container truck path planning control method and device, equipment and a medium, and relates to the field of path planning control, and the method comprises the steps: decomposing a path planning optimization problem into a plurality of independent time domain sub-blocks based on a time domain decomposition method, and obtaining a plurality of control periods of a prediction time domain; for each control period, based on the vehicle dynamics model, the state evolution equation and the operation state data, the current state and control input of the intelligent container truck are updated, and the updated state quantity and the updated control input are obtained; establishing a path planning sub-model of each time domain sub-block based on a Lagrange function, and updating the Lagrange dual variables to obtain updated dual variables; calculating a block moment according to the updated dual variable and the updated state quantity; determining whether a stop criterion is satisfied based on the block moment and the path performance index; and if yes, determining the current target trajectory as a final path planning result, and improving the calculation efficiency and the control precision.

This invention provides a three-trailer collaborative aircraft transfer method and system based on multi-sensor fusion. This invention relates to the field of aircraft ground support technology, specifically to a three-trailer collaborative aircraft transfer method based on multi-sensor fusion, applicable to short-haul aircraft ground transport to hangars. Through multi-sensor data fusion and multi-trailer collaborative control, it improves transfer accuracy, safety, and efficiency. The method specifically includes the following steps: system setup and sensor calibration; trailer attitude initialization and collaborative parameter configuration; real-time acquisition and preprocessing of multi-sensor data; multi-sensor data fusion and state estimation; dynamic path planning based on fused data; attitude error quantification and calibration based on complex statistics; multi-trailer collaborative driving control and traction distribution; real-time environmental monitoring and risk warning; precise docking at the hangar entrance and attitude fine-tuning; and parking and system review and optimization.

The present application relates to the technical field of flexible job shop scheduling in intelligent manufacturing and production scheduling, and particularly relates to an optimization method for solving flexible job shop scheduling with adjusted resource constraints. The method comprises the following steps: initializing parameters and randomly generating an initial population; using a crossover operator and a mutation operator in sequence to evolve the current population; executing a hybrid decoding strategy on the current population; sorting the individuals in the current population in ascending order of maximum completion time to form an elite population, and updating the elite population through problem-specific local search; judging whether the evolution condition is met, and if yes, executing mathematical evolution based on CP, and outputting a final solution when the running time reaches the total running time. The present application has the positive effects of reducing resource waiting time, improving machine utilization, and improving the resource utilization efficiency and scheduling performance of the entire workshop production.

The invention relates to the technical field of AGV monitoring, in particular to an AGV single vehicle state monitoring method based on operation state analysis, and the method comprises the steps: calculating a multi-dimensional operation state index based on the multi-source sensor data of an AGV, and generating a comprehensive state score; dynamically adjusting operation parameters of the AGV monitoring system based on the comprehensive state score and the change trend thereof; and based on the link quality index calculated in real time and the comprehensive state score, cooperatively adjusting a data transmission strategy between the AGV and the cloud. When the communication quality is lower than a preset threshold value, optimizing a local state record according to the comprehensive state score; and carrying out grading diagnosis and verification on the detected abnormity in combination with the comprehensive state score.

This invention discloses a spatiotemporal joint anti-interference method in a UAV-ground cooperative network. The method includes: Step 1: Constructing a UAV-ground cooperative network communication scenario; Step 2: Establishing a base station status discrimination mechanism; Step 3: Constructing a satellite link switching model; Step 4: Setting an anti-ping-pong mechanism for heterogeneous link switching; Step 5: Analyzing the communication effectiveness of heterogeneous links; Step 6: Constructing a joint optimization problem; Step 7: Designing an intelligent anti-interference decision-making method based on DDQN; Step 8: Executing the anti-interference decision. This application autonomously optimizes anti-interference strategies and flight trajectories in complex scenarios involving intelligent interference and base station failures, achieving global optimization of the flight path while ensuring the reliability of the communication link.

The application discloses an active defense method, device and equipment for diffusion speech conversion and a medium, and relates to the technical field of speech security. The active defense method comprises the following steps: obtaining source speech and reference speech, introducing a constrained protection disturbance into the reference speech, constructing protected speech, and ensuring that the protection disturbance satisfies a disturbance imperceptibility constraint. The source speech and the protected speech are input into a speech conversion model with the reference speech as a speaker condition, content information analysis, speaker condition constraint acoustic generation and waveform synthesis are completed by the speech conversion model, and a protection generated speech is output. A joint loss function is constructed based on speaker embedding representations of the protection generated speech and the reference speech, gradients of a current disturbance position and a predicted midpoint position are calculated according to the joint loss function and are fused to obtain an update direction, the protection disturbance is iteratively optimized, and projection or clipping is performed to satisfy the disturbance imperceptibility constraint. Finally, the protected speech is output.

The application relates to an overhead power transmission line inspection robot tower intelligent control method and system, which monitors the real-time residual power of an inspection robot, acquires an environment perception data set to calculate an environment compensation coefficient. A dynamic power threshold is calculated based on the environment compensation coefficient, and a charging request instruction is generated when the power is lower than the threshold. In response to the charging request, the path value of each charging dock is calculated according to the state of each charging dock and the environment data, and the target charging dock and the optimal moving path are determined accordingly. The robot is controlled to move along the path, and the walking parameters are adjusted in real time based on the environment perception data. After the robot enters the docking range, the pose deviation between the robot and the charging dock interface is fine-tuned until the tolerance is met, and a docking success signal is generated. Then, a charging connection is established, the charging current is dynamically adjusted according to the real-time battery temperature and the environment data, and the state data of the target charging dock is updated after charging, thereby improving the safety and reliability of the autonomous charging process of the inspection robot.

This invention relates to the field of intelligent battery swapping recommendation technology, and discloses an intelligent battery swapping recommendation method and system based on historical battery data. The method includes: a vehicle central control system identifying the vehicle from a user-triggered battery swapping request signal to obtain parameter information of the current vehicle; querying historical battery matching data corresponding to the parameter information from a cloud database, and calculating the compatibility score between each battery model and the current vehicle; a battery swapping cabinet control system performing multi-objective collaborative recommendation decision-making with the compatibility score and the inventory batteries to obtain the optimal recommended battery identifier; and the battery swapping cabinet opening the corresponding compartment door according to the optimal recommended battery identifier to execute the battery swapping process and generate a battery swapping operation record. This invention improves the accuracy of battery compatibility assessment, can discover performance correlations between different battery models, and thus improves the automation and intelligence level of the entire battery swapping process.

The invention relates to the technical field of big data processing and intelligent recommendation, in particular to an intelligent recommendation system driven by big data. The method comprises the following steps: collecting user behavior data and article feature data, and generating a user feature set and an article feature set; a user feature weight is introduced, a relative deviation degree is calculated, a recommendation value is calculated, and a recommendation list is generated; displaying the recommendation list to the user to obtain a feedback value; and dynamically adjusting the user feature weight based on the feedback value, the recommendation value and the relative deviation degree, calculating the relative deviation degree and the recommendation value based on the obtained final user feature weight, and finally generating a personalized recommendation list. The problem that a traditional recommendation system mainly depends on static similarity calculation of users and articles, so that personalized recommendation accuracy is insufficient is solved; the response capability to the user interest change is weak, so that the recommendation system easily misjudges the user demand; and a global performance optimization and regulation and control mechanism is lacked, so that a recommendation system is easy to fall into local optimum.

The present application provides a ladle scheduling method and device based on a heat plan, equipment and storage medium, the method comprises the following steps: obtaining the heat plan and the ladle information; formulating the corresponding ladle plan according to the heat plan; determining the planned start time and the planned end time of the ladle plan based on the time information corresponding to the multiple process stages of the heat plan; establishing a ladle matching model according to the ladle plan and the ladle information; taking the highest ladle utilization rate and the least number of matched new ladles as the optimization target; solving the ladle matching model to obtain the matching result of the ladle and the ladle plan; and matching the heat plan and the ladle based on the matching result and the corresponding relationship between the ladle plan and the heat plan, so as to schedule the ladle based on the heat plan. The method solves the technical problems of low efficiency and easy errors in manual ladle scheduling.

The application relates to a large language model assisted cross-layer optical network routing optimization method. The method comprises the following steps: constructing the architecture of the cross-layer optical network, collecting network physical node composition, link resource parameters and service request characteristics; designing an auxiliary graph model, converting the cross-layer network routing optimization problem into a minimum cost path calculation problem in a virtual topology; building a large language model enabled auxiliary graph model weight optimization framework, and preliminarily generating the edge weight of the auxiliary graph model by the large language model; calculating the minimum cost path based on the initial edge weight of the auxiliary graph, evaluating the network performance index corresponding to the path, inputting the evaluation result into the large language model as feedback data, iteratively adjusting the edge weight until the performance index converges to the optimum; and completing the routing planning and wavelength allocation of all service requests in the auxiliary graph model according to the optimal edge weight. The application combines the large language model and the auxiliary graph model framework, improves the flexibility and multi-objective adaptation capability of the routing optimization, and realizes global optimization.

The invention relates to the technical field of storage intelligent management systems, in particular to a cold chain storage intelligent management system based on the internet of things, which comprises an internet of things network constructed by a plurality of cold chain warehouses, and is characterized in that each cold chain warehouse is internally provided with a temperature adjusting device, an image acquisition device, a temperature acquisition device and a storage and transportation planning module; determining a circulation coefficient according to the space utilization rate of the corresponding cold chain warehouse, determining a to-be-circulated storage object list based on the circulation coefficient and the historical warehouse-out data, and responding to the to-be-circulated storage object list or the to-be-stored object list to obtain the to-be-delivered storage object list so as to match the cold chain warehouse. And a warehousing strategy is determined according to the number of the cold chain warehouses which can be matched. According to the invention, the technical problem that the storage environment of the stored objects is not matched with the actual demand due to the change of the storage environment of the stored objects and the space proportion caused by the fact that the warehouse-in and warehouse-out link of the goods is not comprehensively considered in the prior art is solved.

The energy-saving control system comprises a hardware layer and an algorithm layer, the hardware layer comprises executing mechanisms such as a compressor unit, an evaporator and a condenser, a sensor module and a PLC, and the algorithm layer comprises an upper computer algorithm module; according to the energy-saving control method, system operation data are collected through the sensor module and input into the upper computer algorithm module after being preprocessed, local agent decision making is achieved through multi-agent reinforcement learning MARL, global collaborative optimization is completed in combination with a collaborative genetic algorithm CGA, decision consistency is guaranteed through conflict detection and a coordination mechanism, and energy-saving control is achieved. And finally, the PLC controller executes the control instruction and realizes closed-loop feedback updating. The temperature stability, the energy efficiency ratio and the equipment reliability are comprehensively considered, global optimal control is achieved, energy consumption is effectively reduced, the overall operation efficiency of the refrigeration house group is improved, and the method is suitable for the fields of food, medicine, industrial cold chain storage and transportation and the like.

The invention relates to the technical field of power transmission line planning, in particular to a power transmission line path selection system and method based on a database, and the system comprises a data collection and preprocessing module, a data storage and management module, an algorithm module, an application module and a user interface module. And the data acquisition and preprocessing module is used for acquiring multi-source data of satellite remote sensing data, oblique photography data, laser point cloud data, manual inspection data, meteorological data and geological data. According to the method, the power transmission line path is comprehensively evaluated and optimally selected by constructing the standardized database of multi-source data fusion and combining an intelligent optimization algorithm, so that the optimal path planning scheme can be quickly generated according to different constraint conditions and actual requirements, and dynamic adjustment is realized when a power grid fails or the environment is changed; the global optimization and dynamic adjustment of path planning are realized, and the defects of poor data intercommunity, one-sided evaluation and insufficient adaptability in the prior art are overcome.

The deep learning-based video multi-level adaptive compression cross-layer transmission optimization method and system of the application relates to the technical field of video transmission. By designing a compression level selection model based on deep learning, the optimal compression level is output based on the original video data and network environment parameters, and the corresponding compression strategy of the optimal compression level is obtained. A transmission strategy selection model is designed to generate a transmission strategy based on video quality parameters and network environment parameters. A global utility function is defined to calculate the global utility value when the compression strategy and the transmission strategy are assumed to be executed. An optimization objective function is designed, which inputs the compression strategy and the transmission strategy and outputs the optimized compression strategy and the transmission strategy. A condition judgment logic is set to select the final strategy based on the compression strategy and the transmission strategy before and after optimization, thereby achieving global optimization of video compression and transmission.

The present application relates to dilution refrigerator temperature calibration technical field, and disclose a kind of dilution refrigerator thermometer in situ calibration system based on natural cooling dynamics, comprising: according to the dynamics equation of cooling power and temperature Curvature is calculated, according to the temperature corresponding to thermometer resistance value, the temperature is substituted into temperature curvature, and the calibration function of resistance value and temperature is obtained;Identify the multivalued region of temperature, and evolve multiple branch functions in multivalued region in calibration function, calculate the deviation measure of each branch function and calibration function;In multivalued region, the support degree of each branch function to calibration function is calculated, and real-time data is obtained by adjusting calculation load, to minimize the number of experiments to verify calibration function;According to the branch verification result of multivalued region, dynamically adjust the verification strategy of non-multivalued region, under the premise of not relaxing calibration accuracy, through dynamics branch and load guide, significantly reduce the repeated experiment cost of dilution refrigerator cryogenic calibration.

The invention discloses a nucleation monitoring method and system for metal powder, and relates to the field of material preparation.The method comprises the steps that inert gas pressure information and electrode parameter information of a generation chamber are obtained within a first duration, and the electrode parameter information comprises arc current information and arc voltage information; temperature gradient information of the condensation chamber is obtained; and inputting the inert gas pressure information, the electrode parameter information, the temperature gradient information and the temperature gradient information into a nucleation rate model to obtain a nucleation monitoring result.

The invention discloses a method and system for optimizing two-dimensional argon plasma uniformity by introducing a high-energy particle source, and the method comprises the steps: S1, filling a plasma reaction chamber with background gas, and simulating the distribution of neutral particles; s2, initializing the speed and the position of charged particles in the gas in the chamber, enabling the charged particles to be uniformly distributed in the chamber, and determining the intensity, the time sequence, the incident direction and the incident position of high-energy particles emitted by an external particle source; s3, calculation of charged particle density, potential and electric field; s4, calculating the force acting on the charged particles; s5, moving the charged particles; s6, checking and executing collision; and S7, repeating the steps S3 to S6, adjusting the high-energy particle injection time sequence until a steady state is reached, and obtaining the data distribution of the charged particles. Through controllable interaction between an external particle source and plasma or background gas, seed electrons-ions are actively generated in a specific space area or energy distribution is changed, and inherent space non-uniformity of the plasma is compensated.

This invention discloses a method for adjusting the end effector stiffness of a hybrid robot used in the machining of thin-walled structural parts. The method first collects the proportional gain of each active branch servo drive system as the axial servo stiffness; secondly, it establishes a mapping model based on the robot configuration to solve for the equivalent axial servo stiffness of the end effector in the reference coordinate system; subsequently, it introduces stiffness weighting coefficients and combines them with the mechanical stiffness and electromechanical coupling stiffness performance evaluation index; finally, it uses a one-dimensional discrete search algorithm to traverse and filter redundant rotation angles that satisfy displacement and rotation constraints while keeping the end effector pose unchanged, aiming to maximize the electromechanical coupling stiffness index and generate an optimal sequence of redundant rotation angles. This invention establishes an electromechanical coupling model integrating mechanical and servo systems and adaptively adjusts the end effector stiffness using redundant degrees of freedom, which can effectively suppress cutting deformation and chatter during the machining of thin-walled parts while reducing computational complexity, significantly improving machining quality.

The invention discloses a pure highland barley powder extrusion puffing process optimization method, and belongs to the technical field of food extrusion processing. According to the method, the technical problems that the high-fiber highland barley powder is easily subjected to insufficient puffing and too hard texture in the extrusion process, and the retention rate of heat-sensitive functional components such as beta-glucan is low are solved. The method comprises the following steps: optimizing process parameters based on a multi-target response surface method, and determining the water content of materials, the temperatures of a first region, a second region, a third region and a fourth region of a machine barrel and the rotating speed of a screw; carrying out low-temperature superfine grinding on the highland barley powder; a compound emulsifier and edible sodium bicarbonate are added during hardening and tempering; during extrusion, the temperature gradient between the third area and the fourth area is controlled to be smaller than or equal to 20 DEG C; the method is mainly used for producing the high-quality pure highland barley puffed food, the puffing rate and the beta-glucan retention rate of the product can be synergistically improved, the hardness of the product is effectively controlled, and the eating quality and the nutritional value are improved.

The invention provides a federal learning method, a medical data processing method, a medical data processing system and a related device. The method comprises the following steps: acquiring a prediction label carrying a medical sample identifier sent by each participant and a prediction deviation between the prediction label and an original real label; for each medical sample, under the condition that the medical sample is detected to be an abnormal medical sample according to the prediction labels of all participants and the prediction deviation, generating a target real label of the medical sample according to the prediction labels of all participants; and sending a target real label carrying the medical sample identifier to each participant, so that each participant updates a local medical large model according to the target real label corresponding to the medical sample identifier and the prediction label. According to the method, collaborative correction and global optimization of the abnormal medical samples are realized, the interference of the abnormal medical samples on model training is avoided, and the generalization ability and accuracy of the finally generated model are improved.