692results about How to "Improve rationality" patented technology

The invention discloses an agent task scheduling planning method, and relates to the technical field of agent scheduling. The method comprises the steps of analyzing a task instruction to generate atomic tasks capable of being independently executed, constructing a subtask dependency graph, and defining association constraints between the tasks; determining the real-time resource occupancy state of the intelligent agent to obtain a resource state tensor, completing resource-task association anchoring and dependency priority ranking in combination with the sub-task dependency graph, and generating a task priority sequence with resource constraint; performing dynamic capability matching and predictive load balancing calculation on the sequence through a task-agent adaptation model, and determining a target execution agent of each atomic task; and based on the target execution agent and the subtask dependency graph, performing time sequence scheduling arrangement and conflict resolution, and generating a collaborative execution scheme. The method improves the reasonability and efficiency of agent task scheduling, reduces resource conflicts and execution timeout risks, and is suitable for agent cluster collaborative scheduling in a complex scene.

The invention discloses a user track privacy protection method and system fusing a generative model and three-stage personalized differential privacy, and relates to the technical field of data security, and the method comprises the steps: obtaining real user track data, and carrying out the preprocessing of the real user track data to form standardized track representation; training a depth generation model to generate synthetic trajectory data; performing density clustering on the synthetic trajectory to obtain a feature region and calculating a region centroid; point-level privacy budgets are distributed according to three stages of the area proportion, the centroid distance and the global density; constructing an adaptive elliptical disturbance domain based on the regional covariance, sampling candidate confusion points in the disturbance domain and applying road feasibility screening; and calculating a score based on the privacy budget and the utility function, and sampling replacement points by adopting an index mechanism to output a privacy protection trajectory. According to the technical scheme, the track structure continuity and the road semantic consistency are remarkably improved under the condition that the differential privacy constraint is met, and the privacy intensity and the data availability are both considered.

The invention discloses an artificial intelligence model training and reasoning management system based on a knowledge graph, and the system comprises the following modules: a data analysis module which is used for collecting multi-source perception data and forming an entity set and an event set; the graph construction module is used for constructing a context knowledge graph based on the entity set and the event set; the context mapping module is used for establishing a reversible mapping relationship between the context knowledge graph and the original multi-source sensing data; the problem analysis module is used for carrying out semantic analysis on a problem input by a user and cutting the problem to obtain a context sub-graph; the reasoning chain generation module is used for generating a reasoning chain meeting time sequence consistency and causal consistency; the graph reasoning module is used for generating graph reasoning embedding based on the improved T-GNN model and outputting a question and answer result in combination with a question and answer generation model; and the feedback updating module is used for updating the context knowledge graph and the model parameters. According to the method, structured modeling and credibility closed-loop management of the question and answer reasoning process are realized.

The invention discloses an exoskeleton robot abnormal working condition automatic response and risk avoiding decision-making method, and relates to the technical field of exoskeleton, and the method comprises the steps: deploying a plurality of types of sensors to synchronously collect environment data, human physiological data and exoskeleton operation data, and carrying out the preprocessing to remove interference; through fusion of abnormal harmonic detection and a finite-state machine, dangerous working conditions such as wind speed abrupt change are identified, and grades are judged; monitoring eye movement and heart rate variability data in real time, quantifying cognitive load and grading; a decision rule base is constructed, and joint locking and other adaptive risk avoiding decisions are generated in combination with working condition types, grades and cognitive loads; the controller drives the execution mechanism to execute a decision, continuously monitors data at a frequency, dynamically adjusts the decision or recovers a normal mode. According to the method, dangerous working conditions are accurately identified through multi-source data fusion, adaptive risk avoiding measures are automatically triggered, a power assisting strategy is dynamically adjusted in combination with cognitive loads, and operation flexibility and physical output are balanced.

The invention relates to a credit distribution method in multi-agent cooperation, and the method comprises the following steps: obtaining joint data of a multi-agent system, the joint data comprising observation information and execution actions of each agent; the joint data are input into a discriminator model, and the discriminator model carries out modeling on an interaction dependency relationship among multiple agents based on an interaction dependency relationship modeling module of an attention mechanism; generating an auxiliary reward signal corresponding to each agent through a discriminator model; generating a fused reward signal based on the auxiliary reward signal and a global reward signal from the environment; and training and updating the multi-agent strategy model by using the fused reward signal so as to optimize the cooperation strategy of the multi-agent system. By introducing a credit distribution mechanism based on a discriminator model, automatic decomposition and optimization generation of individual reward signals are realized, so that the learning efficiency and stability of a multi-agent system in a complex cooperative task are remarkably improved.

The invention relates to the technical field of voice recognition control, in particular to a Bluetooth sound effect voice interaction method and system, and the method comprises the following steps: completing the energy change analysis and grading recognition based on a voice signal collected by a Bluetooth sound, analyzing channel characteristics in combination with recognition characteristics, summarizing a stable trend, adjusting a priority queue of a sound effect control instruction, and completing the voice interaction of the sound effect of the Bluetooth sound. And matching data packets with different structures and optimizing a distribution process to obtain a delay distribution index. According to the method, voice recognition results are identified in a grading manner, sound effect control instruction priorities are distinguished and sorted in combination with channel stability trend parameters, transmission requirements and feedback responses of different queue instructions are distinguished in a hierarchical data packet structure, task scheduling, the data packet structure and time delay feedback form closed-loop dynamic adjustment, and the control efficiency is improved. Reasonable distribution and feedback management in an instruction transmission process are improved, task response timeliness is continuously optimized, and continuous stability and execution consistency under multi-round sound effect voice interaction are enhanced.

The invention relates to the technical field of enterprise motorcade management, in particular to a motorcade dynamic scheduling and energy consumption collaborative optimization method and system based on digital twinning, and the method comprises the steps: building a motorcade digital twinning scheduling cloud platform, and accessing a plurality of motorcade participants cooperatively executing transportation tasks; the method comprises the following steps: carrying out space-time division on a transportation task according to a collaborative scheduling authority, generating and issuing motorcade scheduling blocks, carrying out initial parameter definition on key factors such as vehicle distribution, path planning, load optimization and dynamic road condition response by each participant based on block information, and constructing a plurality of motorcade dynamic scheduling modeling spaces by combining with traversal search of a scheduling parameter library; an improved teaching and learning optimization algorithm module is introduced, a dynamic competition optimization algorithm is embedded to enhance the adaptability, and a motorcade scheduling collaborative optimization algorithm module is formed. Global optimization is carried out in multiple modeling spaces based on the algorithm, an optimal scheduling parameter set is output, and dynamic scheduling and energy consumption collaborative optimization of transportation tasks are achieved.

The invention discloses a multi-source geographic vector data matching method based on topological relevance, which comprises the following steps: collecting multi-source basic geographic vector data to form a preprocessed data set; constructing multi-index data quality evaluation based on local topological features; extracting and classifying spatial topological relation edges; constructing a weighted adjacent matrix of the target map; a neighborhood topology consistency measurement index is calculated, and neighborhood topology consistency quantization and adaptive weight fusion are carried out; establishing a global optimization objective function by taking the confidence coefficient matrix as input, solving a maximum weight bipartite graph matching problem with topology constraints by taking the topology retention rate as a hard constraint condition, and outputting a global optimal matching scheme; and constructing a quantitative evaluation index topology deviation measure based on topology invariant difference, judging whether a matching scheme has a significant topology conflict, and performing hierarchical repair according to the conflict condition. According to the method, the reliability and logic self-consistency of multi-source vector data fusion are improved, and the key problem of topological contradiction in the prior art is solved.

The invention relates to the technical field of grinding path optimization, and discloses an abrasive paper grinding path optimization system based on reinforcement learning. According to the system, a grinding task is decomposed through a task analysis module, and a state space is defined by fusing a current position, a historical track and real-time surface roughness through a state construction module. The strategy learning module utilizes a deep reinforcement learning network to carry out iterative training by exploring actions and collecting reward experience in simulation so as to learn an optimization strategy. And the trained network inferes local optimization paths for the sub-regions, and a global grinding path is formed after splicing and smoothing. According to the method, self-adaptive path planning is achieved through autonomous decision-making of reinforcement learning, closed-loop quality control is formed by means of real-time roughness feedback, and the intelligent level of grinding machining and the consistency of workpiece surface quality are improved.

The invention discloses a shift change information intelligent complementation and error early warning system based on a large language model, which belongs to the technical field of medical information processing and comprises a terminal acquisition module, an intelligent distribution complementation module and an early warning control module. Structure and semantic fingerprints are generated for the data table, a shift exchange asset meta model is constructed through fusion similarity clustering merging, and a semantic enhancement model is generated in combination with medical shift exchange specifications to complete information field completion; the intelligent distribution and completion module configures an optimal path with the minimum comprehensive overhead for the acquisition module by relying on a regional dynamic shift information network formed by double sub-networks; the early warning control module is combined with a large language model semantic analysis and rule engine, monitors an acquisition track and identifies information abnormity to execute error early warning, and an early warning result is fed back to the intelligent distribution complementation module, so that dynamic adjustment of an optimal path is realized, and the integrity, accuracy and acquisition efficiency of medical shift information are improved.

The invention provides a Bayesian-based aviation equipment residual life uncertainty prediction method and system. The method comprises the following steps: S1, constructing an observation function equation and a state parameter description equation of residual life of key or important systems, modules and parts of aviation equipment; s2, constructing an agent model; s3, solving joint posterior probability distribution and performing residual life prediction; and S4, according to the prediction probability distribution of the remaining life of the aviation equipment, generating an optimization control instruction or a maintenance instruction for the aviation equipment, and acting the optimization control instruction or the maintenance instruction on the aviation equipment. The method provided by the invention not only provides a theoretical basis for solving the problem of prediction, analysis and prediction of the service life of the aviation equipment, but also clearly defines how information and uncertainty thereof flow and convert between a physical world and a digital world, and provides a brand new thought for prediction of the uncertainty of the residual service life of the aviation equipment.

The application relates to the technical field of data processing, and discloses a data security weight-based dynamic adjustment data trusted space permission control method. The method comprises the following steps: constructing a correlation permission matrix; assigning a basic security weight value to a data object, assigning a credit basic value to a data account, coupling the basic security weight value and the credit basic value to obtain a data security weight value; monitoring the attribute change of the data object and the behavior characteristic change of the data account in real time, updating the data security weight value to obtain an updated permission matrix; receiving an operation request, calling a north-south control permission, and checking the access permission of a target data object; when the north-south control permission is checked, the real-time security weight value is obtained, the east-west control permission in the updated permission matrix is matched, and the target operation permission is confirmed; after the target operation permission is confirmed, an operation record is generated, and the behavior characteristic is updated; and the application can improve the efficiency of dynamically adjusting the data trusted space permission control.

This invention relates to the field of multi-UAV cooperative control and target detection technology, specifically to a multi-UAV cooperative search method for underwater targets, comprising: acquiring a probability distribution heatmap; determining UAV parameters and sensor parameters, and establishing the fitness of a genetic algorithm based on the time taken for multiple UAVs to complete all search tasks, total flight time, weights of high-risk hotspots, and arrival times at each target hotspot; optimizing the genetic algorithm based on the probability distribution heatmap and the fitness to obtain an optimal task allocation sequence; performing Dubins transition and spiral search based on the optimal task allocation sequence, the minimum turning radius of the UAVs, and the maximum allowable distance between adjacent routes to obtain a flight path; and for each UAV, tracking the flight path based on the real-time position of the UAV and the eccentricity of the flight path. This invention can improve the anti-interference capability and efficiency of multi-UAV cooperative target detection.

The invention provides a hybrid switch device, and relates to the technical field of electrical equipment. The hybrid switch device comprises a mechanical switch unit, an electronic switch unit, an isolation switch and a protection element. The electronic switch unit and the mechanical switch unit are connected in parallel, the isolation switch is arranged in a loop where the mechanical switch unit is located in parallel and arranged in a loop where the electronic switch unit is located in series, or the isolation switch is arranged in the mechanical switch unit in series and arranged in a loop where the electronic switch unit is located in series; the protection element is connected in parallel with the electronic switch unit; in the switching-on process, the isolating switch, the electronic switch unit and the mechanical switch unit are sequentially switched on, and the electronic switch unit is switched off; and in the opening process, the electronic switch unit is closed, and the mechanical switch unit, the electronic switch unit and the isolating switch are sequentially disconnected, so that the isolating switch ensures the electrical isolation safety, and the reliability, controllability and connection layout rationality of the hybrid switching device are improved.

The application discloses a large and medium-sized water and electricity engineering emergency resource assessment and prediction method and device, relates to the safety emergency technical field of water conservancy and hydropower engineering, and specifically discloses the following technical scheme: multi-source heterogeneous data such as engineering basic information, emergency resource account books and expert knowledge rules are collected, a standardized data base is constructed through cleaning and checking and semantic mapping, an evaluation index system is constructed based on the data base, the emergency resource guarantee capacity is quantitatively evaluated and the short board is identified through a comprehensive evaluation algorithm, the evaluation result is combined with an emergency scene, a task chain is deduced, resource demand is calculated and compared with inventory, the optimal storage quota is determined through multi-scenario weighted optimization, finally, a multi-target model is constructed in combination with a budget and material priority, a phased scheme is intelligently optimized and generated, visualized interaction is output, and the optimal allocation and efficient scheduling of emergency resources are completed. Through multi-target intelligent optimization and visualized interaction, the phased scheme output is realized, and the problems of slow emergency response and unscientific decision-making are solved.

This invention discloses a method for constraining corner elements in a rocking wall structure system. Based on the structural parameters of the rocking wall and the frame structure, the shear stiffness of the frame structure is obtained. A first lateral displacement expression and a second lateral displacement expression are constructed using the shear stiffness and the first and second equilibrium equations. An inter-story drift angle expression is then constructed using these expressions to solve for the stiffness of the rotating combined disc springs. The bending moment of the rotating combined disc springs is obtained from the stiffness of the rotating combined disc springs. The number of single disc springs stacked in the same direction and the number of combined disc springs are determined by the deformation requirement and the bearing capacity requirement of the combined disc springs obtained from the bending moment. This yields the combined height of the disc springs. The combined height of the disc springs is then used to constrain the length of the core rod and the support height in the corner element, achieving rapid constraint of the core rod length and support height in the corner element based on statics.

The invention relates to the technical field of virtual geographic scene visualization, and discloses a virtual geographic scene enhanced visualization method, device and equipment and a storage medium, and the method comprises the steps: obtaining a virtual geographic scene visualization request of a user terminal, considering the visualization constraint condition provided by the user terminal, and improving the visualization efficiency of the virtual geographic scene. The method comprises the following steps: rendering and mapping geographic information data into a visual display scene, constructing a scene visual feature set, and according to a selective visual expression demand of a user terminal for each scene visual feature; the scene visual features in each visual display period are subjected to display content selective expression, the scene display content with the highest visual cognition efficiency is sent to the user terminal, and finally the scene display content of the next visual display period is adjusted according to the display evaluation result fed back by the user terminal. While the operation load is reduced, the scene visualization experience of the terminal user is improved, and the scene display strategy which occupies the least system resources and can meet the current user demand is provided.

The present application relates to the technical field of vehicle pseudonym replacement, and particularly relates to a vehicle pseudonym replacement method based on artificial intelligence, which comprises the following steps: determining whether pseudonym replacement is needed according to a replacement time interval value or performing secondary determination based on a quantity ratio reference value and a communication characteristic value; determining a target area type according to a to-be-replaced vehicle distribution reference value and a replacement allowance, and determining whether to directly perform pseudonym replacement or perform combined analysis according to the target area type; in the combined analysis, determining a related vehicle combination based on a position synchronization degree or an interaction reference value according to a vehicle analysis coefficient, determining whether to perform sequential setting of a virtual vehicle based on a quantity difference value or setting a virtual vehicle based on a historical attack frequency of a sub-area, and performing pseudonym replacement on to-be-replaced vehicles and virtual vehicles in the related vehicle combination. The present application can reduce the risk of vehicle location privacy leakage.

This invention provides a method and system for UAV path planning based on a fusion algorithm, relating to the field of UAV autonomous navigation and path planning technology. The method includes the following steps: acquiring the starting coordinates, ending coordinates, and obstacle information in the flight environment of a flight mission, and constructing a flight environment spatial model; defining path constraints and optimization objectives, and generating an initial set of feasible paths based on the flight environment spatial model, path constraints, and optimization objectives; performing weighted fusion and guided enhancement on the initial set of feasible paths to obtain a weighted fused path and a guided enhanced path; employing a particle swarm optimization algorithm to construct an initial particle swarm based on the initial feasible path set, the weighted fused path, and the guided enhanced path, and iteratively optimizing the initial particle swarm to obtain a globally optimal particle; and decoding the globally optimal particle to obtain the final flight path. This invention significantly improves the efficiency and quality of UAV path planning in complex static obstacle environments.

This utility model relates to vehicles, specifically providing a differential, a vehicle transmission assembly, and a vehicle, aiming to solve the problem of excessive internal oil pressure affecting the sealing effect of the oil seal in the differential. To this end, the differential of this utility model includes: a support base with a first through hole, the inner wall of which has an annular groove; an input shaft; and an adjusting ring disposed within the annular groove, with a clearance fit between the adjusting ring and the inner wall of the annular groove, allowing oil to flow through the gap on both sides of the adjusting ring. This utility model, by setting an adjusting ring within the through hole of the differential support base and ensuring a clearance fit between it and the inner wall of the annular groove, achieves free flow of oil under normal operating conditions; when a large oil pressure difference occurs on both sides of the adjusting ring, the adjusting ring automatically adheres to the inner wall of the annular groove under pressure, promptly sealing the gap and effectively isolating the oil flow. This structure can dynamically adjust the internal oil pressure of the differential without changing the original arrangement.

The invention discloses an inspection method, device and equipment for an oil and gas exploitation system and a storage medium, and belongs to the technical field of oil and gas exploitation. The method comprises the steps that basic information associated with an inspection point is acquired, the inspection point comprises a to-be-inspected equipment assembly in the oil and gas exploitation system, and the basic information represents the association relationship between the inspection point and inspection equipment; generating first inspection information by using the basic information and the inspection task of the inspection point; performing conflict detection on the first inspection information, and optimizing the first inspection information based on a conflict detection result to obtain second inspection information; routing inspection is conducted on the routing inspection point based on the second routing inspection information, first routing inspection data and second routing inspection data are obtained, the first routing inspection data comprise routing inspection images collected by the image collection device, and the second routing inspection data comprise information collected by the device sensor; and generating an inspection result based on the first inspection data and the second inspection data. The method can improve the inspection accuracy and inspection efficiency of the oil and gas exploitation equipment in the oil and gas exploitation system.

The application discloses a kind of optimization methods of tubular heat exchanger low-carbon manufacturing process parameters, it is related to tubular heat exchanger manufacturing technical field, including the following steps: according to process, tubular heat exchanger manufacturing process is split, is divided into multiple with carbon emission related process object, the corresponding relationship between process parameters and carbon emission is established, and the before-and-after relationship and parallel relationship between each process are described using AND / OR logic, form process association relationship.The application integrates each process carbon emission calculation by establishing multilevel mapping relationship, realizes whole-process low-carbon collaborative optimization;Combining Sobol method identifies key process parameters, improves parameter screening efficiency and rationality.At the same time, using double population and fuzzy logic regulation mechanism improves optimization ability under complex constraint, the obtained parameter can be directly used for production, realizes the comprehensive optimization of carbon emission, processing time and cost without equipment modification.

This application discloses an interaction method, device, equipment, and medium for a card-based game robot, relating to the field of computer technology. The game robot is a virtual game character controlled by an artificial intelligence algorithm. The method includes: training a supervised learning model using PyTorch by parsing real game logs to obtain a target supervised learning model; training a reinforcement learning model, including a card-playing model and an action decision model, based on a distributed model-free control architecture and logs in a simulated environment to obtain a target reinforcement learning model; in a target game, determining whether it belongs to a preset extreme game based on the current game state information and the hand prediction rules in the current game robot; and determining the current game strategy based on the judgment result, the target supervised learning model, the target reinforcement learning model, and the strategy fusion rules in the current game robot, in order to interact with the player character in the game. This application can improve the anthropomorphism and rationality of the game robot's game.

The application discloses a kind of based on artificial intelligence's comprehensive pipe gallery intelligent decision system and method, belong to intelligent decision-making technical field.The application is aimed at each pipe gallery area, pre-calibrates ontology parameter and water seepage boundary parameter, real-time collection operation data, after pre-processing, form comprehensive pipe gallery data.Calculate static water head, combine cabin inside and outside air pressure difference correction to obtain effective action water head;Using artificial intelligence algorithm to predict the change process of ground water level, generate prediction water inflow time sequence.Matching water level and volume corresponding relationship, calculate pipe water total inventory sequence, calculate time period water net increment, combined with effective water cross-sectional area to obtain next minute end prediction water level and water level rising rate;Combined with drainage equipment performance to determine the risk of water inflow in the region.Determine the risk of water inflow, preferentially start local drainage equipment full load operation, as needed control regional interval switch opening degree, linkage adjacent pipe gallery area with drainage redundancy collaborative drainage.

This invention belongs to the field of robotic vacuum cleaner technology, specifically relating to an adaptive cleaning method and system for robotic vacuum cleaners based on debris recognition. The invention includes generating a deviation cleaning path from an initial cleaning path. Specifically, this involves acquiring images of debris within the cleaning area, determining the distribution location information of the debris based on the images, generating an initial cleaning path based on the distribution location information, identifying the intersection area between the initial cleaning path and the debris distribution location information, and generating a deviation cleaning path to cover the debris corresponding to the intersection area. By generating a deviation cleaning path, this invention achieves precise cleaning of the debris's distribution location information, thereby improving the cleaning task coverage and ensuring thorough cleaning of the debris area. Simultaneously, the closed-loop optimization process established in this invention enables the robotic vacuum cleaner to possess adaptive learning and continuous optimization capabilities, enhancing its autonomy and robustness in complex environments.

The invention discloses a waveform feature extraction method and system for irregular waveforms of a hyperspectral laser radar, and aims at solving the problem that asymmetric waveforms, weak echoes and overlapped echoes in multi-channel full-waveform laser radar signals are difficult to accurately decompose, modeling full-waveform signals collected by multiple spectral channels into superposition of echo components and background noise. A multi-channel waveform is preprocessed through noise threshold and pulse width combined constraint, the waveform position is determined by adopting a gravity center discrimination method, and time synchronization calibration of each channel signal is realized in combination with a multi-echo quality index. The method further utilizes a cubic B-spline function to carry out modeling on transmitted waveforms of all spectrum channels, carries out weighted accumulation on multi-channel echo signals to enhance weak echo signals, and adopts a multi-channel interconnection decomposition optimization strategy to realize accurate extraction of irregular waveform features under the constraint condition of keeping time parameters of sub-echoes of all the channels consistent.

The application discloses a kind of mineral resources intelligent prediction method based on PINNs, it is related to mineral resources exploration technical field, including, acquisition mining area multimodal geoscience data, and standardization pretreatment is carried out;Based on mineral resources probability distribution map, spatial target area is automatically delineated, and output hierarchical target area sequence, based on hierarchical target area sequence and mineral resources probability distribution map, application geostatistics method simulates the distribution of different confidence under resource grade and tonnage, and outputs target area resource quantity evaluation report;Uncertainty quantification and visual display are carried out to hierarchical target area sequence and target area resource quantity evaluation report, and comprehensive prediction result drawing is generated.The application realizes the semantic level alignment of multi-source heterogeneous data by constructing geological ontology knowledge graph and using graph convolution network to learn geological concept embedding, and effectively captures the deep feature association related to mineralization in multi-source heterogeneous data.

This application provides a display device and a household appliance. The display device includes a first display section and a knob assembly. The first display section extends along a first direction. The knob assembly is arranged adjacent to the first display section in the first direction. The knob assembly includes a knob operating section and a second display section disposed radially inside the knob operating section. The dimension of the first display section in a second direction is not less than the dimension of the second display section in the second direction and not greater than the outer diameter of the knob operating section. The first and second directions are perpendicular. In this application embodiment, the display dimension of the second display section in the second direction is between the dimension of the second display section in the second direction and the outer diameter of the knob operating section. This allows the second display section to have a suitable size to match the knob operating section, and the overall size is more ergonomic, which helps to improve the overall layout rationality and visual effect of the display device.

The invention discloses a genome variation cold and hot spot region prediction method and device, and the method comprises the steps: S1, carrying out the slicing of a target genome region according to a preset sliding window length, and constructing a multi-modal input tensor corresponding to a window; s2, inputting the multi-modal input tensor into a pre-trained deep learning prediction model, and outputting a cold and hot spot prediction score of each site in the window through a full connection layer; and S3, according to the cold and hot spot prediction scores, identifying a variation cold spot region and a variation hot spot region in the target genome region. According to the technical scheme provided by the invention, the dependence on the existing variation data density is eliminated, and non-blind area coverage in the whole exon group range is realized; meanwhile, the structured output based on the preset transcript coordinates can directly support clinical variation interpretation, a quantitative basis is provided for PM1 and cold spot evidence, and the proportion of unclear significance variation is effectively reduced.

The embodiment of the invention provides a workflow allocation method and device based on a multi-cloud environment, equipment and a storage medium. The method comprises the following steps: determining a to-be-allocated workflow, wherein the to-be-allocated workflow comprises a plurality of to-be-allocated tasks with an execution time sequence relationship; obtaining a resource state set corresponding to the plurality of cloud service nodes, wherein the resource state set comprises a plurality of virtual machine states corresponding to the plurality of cloud service nodes; generating a target input text according to the to-be-allocated workflow and the resource state set, and inputting the target input text into the large language model to obtain node allocation information; reading a target virtual machine state set of a target cloud service node corresponding to the node allocation information from the resource state set; and inputting the to-be-allocated workflow and the target virtual machine state set into a pre-trained target decision model to obtain a task allocation relationship between the plurality of to-be-allocated tasks and the virtual machines on the target cloud service node. Therefore, the reasonability of workflow allocation can be improved, and the resource utilization efficiency is improved.