95 results about "Parametric simulation" patented technology

The invention relates to the technical field of hemodialysis teaching systems, and discloses a hemodialysis teaching system based on AI intelligent pushing. The system comprises an operation feature extraction module, an instrument positioning analysis module, an operation step analysis module and a teaching strategy generation module. Operating a feature extraction module to obtain a hemodialysis teaching video stream, and extracting multi-modal features containing an instrument movement track sequence and physiological parameter simulation data; an instrument positioning analysis module calls a pre-training network for positioning and generates a coordinate data set containing instrument types and three-dimensional position coding mapping; the operation step analysis module analyzes the time sequence characteristics by means of an action segmentation model, and generates a result containing step integrity and an error operation mark; and the teaching strategy generation module inputs a result into the intelligent push model, plans a personalized teaching path, outputs a knowledge point sequence containing a student ability model and dynamic teaching elements of operation difficulty parameters, and assists the hemodialysis teaching to be precise and personalized.

The invention provides a carbon dioxide injection well perforation parameter optimization method and device and electronic equipment, and the method comprises the steps: carrying out the standardization processing of geological and logging data, and constructing a multi-dimensional data set comprising lithology, porosity, permeability, crack parameters and a ground stress field; a three-dimensional digital twinborn model is constructed based on the data set, and reservoir heterogeneity and fracture network distribution are restored; on the basis of the digital twinborn model, performing multi-physics field coupling simulation including carbon dioxide migration simulation, shaft stress distribution simulation and long-term corrosion prediction; based on a multi-objective optimization algorithm, with the effective wave and volume of the carbon dioxide, the integrity coefficient of the shaft and the anti-channeling safety coefficient as objectives, an optimal perforation parameter combination is screened; and outputting a visual report including perforation parameters, simulation results and risk assessment. Digital twinning, multi-physics field simulation and multi-objective optimization are deeply fused, and a set of complete, accurate and efficient perforation parameter optimization solution from data to decision is formed.

The invention discloses an ice rink re-paving ice layer thickness prediction method, system and equipment based on digital twinning, and relates to the technical field of ice rink ice surface thickness. The method comprises the following steps: step 1, constructing a digital twinborn model of the re-paved ice surface of the ice pouring vehicle through COMSOL numerical simulation software, and mapping and fusing the digital twinborn model of the re-paved ice surface of the ice pouring vehicle and real-time environmental data to generate a dynamic digital twinborn body; 2, real-time operation parameters of the ice pouring vehicle are input into the dynamic digital twin body, operation is simulated, and the simulated icing thickness is output; and 3, inputting the simulated icing thickness, the real-time operation parameters of the ice pouring vehicle and the real-time environment data into the trained machine learning model to obtain the thickness of the re-laid ice layer. According to the method, the icing thickness of the re-paved ice surface of the ice pouring vehicle corresponds to the twin model through the digital twin technology, so that the re-paved ice layer thickness prediction is more accurate and comprehensive, the re-paved ice layer thickness prediction is realized, the risk of different ice layer thicknesses is reduced, the waste of water resources is avoided, and the energy consumption is saved.

The invention provides a deep foundation pit monitoring method based on BIM-physical model hybrid simulation. The method comprises the steps that real-time monitoring data and BIM model parameters of a deep foundation pit are collected; inputting a first model parameter in the BIM model parameters as an initial geometric condition into the initial physical model for simulation calculation to obtain first simulation data, and verifying the first simulation data based on the real-time monitoring data to obtain a verification result index; optimizing the soil mass mechanical parameters in the initial physical model based on the verification result index until the verification result index output in the optimization process meets a preset requirement, and obtaining an optimized physical model; simulating the construction process of the deep foundation pit on the basis of the optimized physical model in combination with a second model parameter in the BIM model parameters, and outputting second simulation data; and performing monitoring and early warning on the deep foundation pit based on the deep foundation pit knowledge graph in combination with the real-time monitoring data and the second simulation data to obtain a monitoring and early warning result. According to the invention, comprehensive monitoring of the deep foundation pit is realized.

The invention discloses a thermal-mechanical coupling behavior numerical simulation system for a reinforced concrete structure, and particularly relates to the field of constructional engineering, and the system comprises a heat conduction analysis module, a mechanical behavior analysis module, a bond slip behavior analysis module and a simulation calibration module. The heat conduction analysis module simulates the temperature field change of the concrete in the fire disaster process by defining temperature related parameters of the concrete; the mechanical behavior analysis module is used for simulating nonlinear mechanical behaviors of concrete and analyzing strength degradation, crack propagation and plastic deformation of the concrete at high temperature in combination with a temperature-dependent material model; and the bond slip behavior analysis module is used for setting a non-linear contact behavior between the reinforcing steel bar and the concrete. According to the invention, comprehensive simulation of reinforced concrete under high-temperature conditions such as fire disasters and the like is realized, and the strength degradation and crack propagation of the concrete and the bond slip behavior of the reinforcing steel bars and the concrete can be accurately predicted in combination with a temperature field and a mechanical field.

The invention belongs to the technical field of microstructure simulation in the solidification process, and particularly relates to a nickel-based superalloy laser welding solidification structure evolution law prediction method and system.The method comprises the steps that simulation parameters are determined, a macroscopic temperature field is simulated to obtain molten pool temperature field data, a cellular automaton model is constructed, and grids are divided; initializing state variables such as component concentration and solid-phase fraction; coupling temperature field data, solving an optimized solute diffusion and redistribution equation to obtain a concentration field, updating a solid phase fraction according to local supercooling drive, and marking grain growth; counting a solid-phase fraction, and updating a solid-phase position; and judging whether the temperature of the cladding layer is lower than the solidus temperature and the solid fraction reaches a complete solidification threshold value or not, and if yes, outputting weld microstructure and target element concentration data. According to the method, the evolution law of the welding macroscopic temperature field and the structure and solute concentration field can be effectively predicted, experiments are replaced with numerical simulation, the cost is reduced, the technical blank is filled up, and reliable support is provided for precise regulation and control of the nickel-based high-temperature alloy laser welding process.

The invention discloses an oil and gas development favorable area prediction method based on Winsored particle size mean curve simulation, and the method comprises the following steps: based on laser particle size experiment data, employing a Winsored algorithm to process rock particle size data, removing abnormal interference values, and simulating a well hole particle size mean longitudinal curve through inputting logging response parameters by means of an XLSTM-Transform hybrid neural network model; selecting a shear modulus density and a longitudinal and transverse wave velocity ratio as sensitive attribute parameters of a granularity mean value in combination with seismic data, and forming two sets of three-dimensional attribute data volumes in combination with waveform structure inversion; according to the method, model training is carried out on the nonlinear relation between the particle size mean value and the seismic sensitive attribute parameters based on a machine learning method, and finally a Winsored particle size mean value inversion body is formed, so that the spatial distribution rule of a relatively large particle size reservoir can be defined, and effective prediction of a development favorable area is realized.

The invention discloses a method and a system for determining deformation and warping parameters of an injection molding plate. The method comprises the following steps: acquiring simulation deformation data reflecting the deformation and warping degree of the injection molding plate; judging whether the change trend of the simulation deformation data relative to the standard deformation data is the same and meets a measurement index or not; if the change trends are the same and meet the measurement index, storing the first deformation warping parameter; if the change trends are the same but do not meet the measurement index, modifying a first deformation warping parameter corresponding to the simulation deformation data; if the change trends are different, the simulation deformation data are obtained again after the simulation mode for obtaining the simulation deformation data is switched, and the simulation deformation data are compared with the standard deformation data again. According to the method provided by the invention, the deformation warping of the injection molding plate can be simulated based on the preset deformation warping parameters, the obtained simulation deformation data reflecting the deformation warping degree of the injection molding plate is compared with the standard deformation data, and the corresponding deformation warping parameters are stored when the change trend and the measurement index are met.

The application discloses a mine roadway on-demand support structure optimization method based on artificial intelligence, relates to the technical field of mine roadway support, and collects roadway measurement data of a roadway measurement point; captures and extracts the roadway measurement data to obtain fissure geological parameters, classifies and summarizes the fissure geological parameters to obtain a rock mass quality grade; constructs a roadway section model, collects parameters, obtains original rock mass mechanical parameters, simulates and constructs the roadway section model to obtain a same-grade simulation parameter library; compares and simulates the same-grade simulation parameter library to obtain a parameter simulation result, enhances defects of the roadway measurement point according to the parameter simulation result, and obtains an optimal support scheme; the support scheme design efficiency and accuracy are improved, the roadway stability and safety are improved, excessive support or insufficient support is avoided, and support economy is optimized.

The utility model relates to the technical field of traceability of electric vehicle power battery detection equipment, and specifically provides a parameter simulation-based traceability device for electric vehicle power battery detection equipment. The traceability device comprises traceability equipment and a charging female seat, the traceability equipment comprises a standard voltage source and a control switch, the standard voltage source is electrically connected with the charging female seat through the control switch, the charging female seat is electrically connected with detection equipment, and the detection equipment refers to electric vehicle power battery detection equipment; the standard voltage source is used for simulating the ideal voltage of the power battery of the electric vehicle in the traceability configuration stage; and the detection equipment is used for acquiring the ideal voltage of the power battery of the electric vehicle and performing traceability configuration based on the ideal voltage of the power battery of the electric vehicle and the traceability equipment. According to the utility model, the ideal voltage of the power battery of the electric vehicle is simulated through the standard voltage source, so that the matched connection of the detection equipment and the traceability equipment is realized, and the technical problem that the traceability device cannot simulate the relevant parameters of the power battery of the electric vehicle is solved.

The invention discloses a tilting-pad thrust bearing pad machining method and system. The method comprises the steps that pad basic structure parameters, material physical attribute parameters and preset machining precision indexes are obtained firstly; constructing a tile force thermal coupling deformation simulation model, and inputting parameters to simulate stress distribution and deformation of the tile in cutting; a surface microtopography oil film adaptation model is constructed based on the deformation data, and the mapping relation between the surface microcosmic parameters and the oil film performance is determined; then multi-source sensing equipment is deployed to collect cutting force, temperature and vibration signals, and comprehensive characteristic parameters are obtained through processing of a multi-source sensing cutting force fusion algorithm; the adaptive relation parameters and the comprehensive characteristic parameters are input into a machining parameter decision support platform, and parameters such as the cutting speed and the feeding amount are dynamically optimized; and finally, processing according to the optimized parameters and feeding back data correction parameters in real time. The system comprises six units. According to the method and system, multi-model collaboration and parameter real-time optimization are achieved, and the pad machining precision and the bearing reliability are improved.

This invention provides a three-dimensional finite element non-blasting excavation numerical simulation method for rock mass excavation on waterway slopes. This method utilizes FORTRAN language code throughout the entire process to achieve 3D model mesh creation, excavation calculation, solution, and data visualization. By constructing a digital pre-simulation method using full-process code, it enables a forward-looking assessment of the matching between construction methods and rock mass conditions. Through multi-condition parametric simulation, such as different lithology combinations and excavation force combinations, this method can quantify the response of different construction methods to the rock mass under corresponding conditions. Based on the results, the calculation of parameters such as excavation efficiency and excavation volume can be further expanded.

The application relates to the technical field of hemodialysis teaching systems, and discloses a hemodialysis teaching system based on AI intelligent pushing. The system comprises four modules, namely, an operation feature extraction module, an instrument positioning analysis module, an operation step analysis module and a teaching strategy generation module. The operation feature extraction module acquires a hemodialysis teaching video stream and extracts multi-modal features containing instrument motion trajectory sequences and physiological parameter simulation data; the instrument positioning analysis module calls a pre-trained network for positioning and generates a coordinate data set containing instrument type and three-dimensional position encoding mapping; the operation step analysis module analyzes time sequence features by using a motion segmentation model and generates a result containing step integrity and error operation labels; and the teaching strategy generation module inputs the result into an intelligent pushing model, plans a personalized teaching path, outputs dynamic teaching elements containing a knowledge point sequence of a student ability model and operation difficulty parameters, and helps to realize precise and personalized hemodialysis teaching.

The application discloses a reservoir parameter inversion method based on a proxy model and adaptive iterative pruning, and relates to the technical field of oil and gas reservoir development.The application obtains simulation production data based on prior information of reservoir parameters and a reservoir numerical simulator, generates a plurality of reservoir parameter simulation data to construct a data set, establishes a proxy model based on a convolutional neural network module and a recurrent neural network module, and is used for predicting reservoir production data, sorts weight parameters of each neural network in the proxy model after training and verification of the proxy model by using the data set, adopts an adaptive iterative pruning method, prunes the weight parameters of the proxy model by selecting reservoir parameter simulation data in a database, obtains a pruned proxy model, optimizes a reservoir numerical simulation model based on a set smoother algorithm and the proxy model, obtains an optimal reservoir numerical simulation model, and simulates production dynamic data of an actual reservoir by using the optimal reservoir numerical simulation model, so that the inversion efficiency of the reservoir parameters is effectively improved.

The invention discloses an airport aircraft taxiing efficiency redisk analysis method and system. The method comprises the following steps: performing space-time alignment on flight plan, aircraft position distribution, field monitoring radar and airport notification data to generate a fused space-time database; constructing a weighted directed graph model which takes the airport physical position as a node and the slidable path as an edge as an airport operation map; identifying flights with abnormal taxiing time, and mapping the actual track into an actual taxiing path sequence on the map through a hidden Markov map matching algorithm; calculating a theoretical optimal sliding path sequence by taking starting and ending nodes as constraints; the two paths are compared to position deviation nodes, and delay attribution is carried out in combination with node and edge attributes; and finally, carrying out visual display and parameterized simulation deduction. According to the method and the system provided by the invention, through a data fusion and map modeling technology, automatic and precise root cause tracing and strategy evaluation for the problem of low sliding efficiency are realized, and the intelligent level of airport ground operation management is improved.

This invention relates to the field of submersible electric pump (FET) fault diagnosis technology, and particularly to a method and system for FET fault diagnosis based on CFD and deep learning. The method includes constructing a full-channel 3D model, simulating FET faults by modifying key geometric feature parameters, and using a hybrid mesh strategy to divide the mesh; extracting time-domain and frequency-domain features from the simulation data, forming feature vectors from these features, and constructing multi-dimensional labeled simulation data using these feature vectors and multi-dimensional label vectors; constructing a multi-channel deep neural network, training the multi-channel deep neural network using the multi-dimensional labeled simulation data; and deploying the trained multi-channel deep neural network to a well site edge computing gateway to diagnose FET sensor data. This invention utilizes CFD to generate high-quality data to drive deep learning model training, achieving rapid and accurate collaborative diagnosis of multiple faults in FETs.

The application discloses a deep shale gas reservoir parameter simulation method, device, equipment, medium and product, and relates to the technical field of oil and gas development. The method comprises the following steps: obtaining a preset geological model; simulating a fracturing-production process of the preset geological model based on a preset fracturing-production coupling model; the preset fracturing-production coupling model comprises a wellbore flow model, a fracture flow model and a shale reservoir flow model; iteratively solving based on the preset fracturing-production coupling model and the preset geological model to obtain simulation parameter distribution data in the fracturing-production process; the simulation parameter distribution data comprises one or more of the following: a fracture morphology distribution after fracturing, pressure distribution data of a shale reservoir after fracturing, water saturation distribution data of the shale reservoir after fracturing, pressure distribution data of the shale reservoir after production and water saturation distribution data of the shale reservoir after production. The application can improve the accuracy of deep shale gas reservoir parameter simulation.

The invention specifically discloses a simulation scene generalization method based on Monte Carlo sampling. The simulation scene generalization method comprises the following steps that S1, the minimum generalization number of different distribution functions during generalization is determined; s2, analyzing different generalization parameters to obtain a distribution function most suitable for the parameters, and adopting different distribution functions to meet the challenge of traffic scene diversification; s3, traffic scene parameters conforming to a certain specific distribution function are obtained based on a Monte Carlo sampling algorithm; generalized parameters are generated by using Monte Carlo random sampling, wider and diversified scene changes are simulated, vivid and diversified scene data are generated, and an automatic driving simulation scene with uncertainty is generated by using a scene generation library based on the generated parameters, so that the automatic driving simulation effect is improved. By simulating diversified and vivid driving scene changes, the method can significantly improve the generalization ability and adaptability of the model in different scenes.

The application discloses a tunnel electromechanical intelligent control simulation teaching system and method and a practical training platform, the method comprises a tunnel simulation module, a sensing simulation module, simulated electromechanical equipment and a control module, the tunnel simulation module is used for simulating a real tunnel structure; the sensing simulation module is used for simulating and collecting at least one environmental simulation parameter in the simulated tunnel; the simulated electromechanical equipment is used for simulating real electromechanical equipment in the tunnel; the control module is in communication connection with the sensing simulation module and the simulated electromechanical equipment respectively, the control module is configured to generate linkage control instructions for driving at least two types of simulated electromechanical equipment in the simulated electromechanical equipment according to at least one environmental parameter collected by the sensing simulation module; according to the linkage control instructions, the at least two types of simulated electromechanical equipment are controlled to run, so as to simulate the running state of the electromechanical equipment of the real tunnel. The conversion from textbook theory to hands-on operation is realized, it is guaranteed that students are no longer bystanders, and the learning immersion is greatly improved.

The invention belongs to the technical field of energy exploitation, and particularly relates to a multi-objective optimization method and system for gas reservoir injection, extraction and storage, and the method comprises the steps: firstly obtaining regulation and control parameter variables of extraction and storage, and calling an improved sampling algorithm for sampling combination to generate sample data; calling CMG data files for batch modification, extracting well names and DATE commands from the original dat file, simulating by combining sample parameters, generating a new scheduling block, inserting the new scheduling block into the last DATE command of the original file, generating a new CMG file, and exporting and merging CSV after processing; constructing a multi-output regression prediction model, and training the model by using a CSV file; then all regulation and control parameter variable combinations are extracted, data are exported, the trained model is called for prediction, and a parameter full combination is obtained; and finally, constructing a multi-objective optimization model, and exporting a Pareto optimal solution as a result by taking maximization of cumulative yield and storage efficiency as an objective. According to the invention, the problems of low efficiency and inaccurate precision of collaborative operation of sealing and natural gas exploitation in the prior art can be solved.

This invention relates to the field of traffic safety technology, specifically to an efficient simulation method for quantum systems based on parallel order reduction. The method includes the following steps: receiving the physical parameters of the quantum system and simulation requirements, whereby the physical parameters include electron mass, potential well size, and initial wave packet parameters; and the simulation requirements include simulation physical time and accuracy requirements. Based on the physical parameters, a time-dependent Schrödinger equation describing the dynamic behavior of the quantum system is established, and the Schrödinger equation is rearranged into matrix form using the Kronecker product. This invention utilizes Arnoldi to reduce the order of the matrix-form Schrödinger equation. By reducing the order, the large matrix in the original space is projected into a relatively small subspace, improving the efficiency of the simulation solution. The reduced-order Schrödinger equation is solved using a time-parallel algorithm. This approach overcomes the time step limitation imposed by the CFL condition, ensuring that a stable solution can be obtained with fewer time steps.

The application relates to an enterprise operation optimization method and system based on AI and digital twinning. The operation optimization method comprises the following steps: collecting historical data and real-time data streams in an enterprise business system and processing the same to generate a structured data set; calling a machine learning model to perform feature analysis on the structured data set and output a key indicator prediction result set; constructing a digital twinning virtual factory model and embedding the key indicator prediction result set in the digital twinning virtual factory model; receiving order demand parameters input by a user, simulating an order execution process, and generating an optimization scheme feasibility report; if the optimization scheme feasibility report is verified, converting the optimization scheme into an execution instruction set and pushing the same to a corresponding business system for execution; and capturing execution data fed back by the business system in real time and updating the machine learning model and the digital twinning virtual factory model based on a deviation feedback value of the execution data. The application realizes accurate perception, scientific prediction and dynamic regulation and control of the whole enterprise operation process.

The application discloses a micro-grid building simulation training system based on a block chain, and relates to the technical fields of block chain application and power system simulation training. The system comprises a request analysis module, a simulation kernel module, a data encapsulation module and a consensus writing module. The request analysis module analyzes initial configuration parameters of a micro-grid topology structure contained in a simulation training request initiated by a student terminal. The simulation kernel module generates a virtual micro-grid component instance set based on the parameters and calls a discrete event simulation kernel pre-deployed in a block chain smart contract. The data encapsulation module encapsulates electrical attribute parameters of each component instance into a to-be-verified state data block. The consensus writing module performs hash collision verification on the to-be-verified state data block through a block chain consensus node, packs and writes data blocks that pass the verification into a private data channel of a block chain ledger. The step driving module drives the discrete event simulation kernel to step and update running state parameters of each component instance according to the written data blocks and a preset time axis.

The invention relates to the technical field of new energy automobile charging detection and calibration, in particular to a dynamic parameter simulation new energy charging detection and calibration system, which comprises a dynamic parameter simulation module, an electrical output module, a communication simulation module, an environment acquisition module, an error fusion module and a calibration strategy module. The dynamic parameter simulation module generates a time-varying target parameter time sequence according to a plurality of continuous stages in a preset calibration scene and outputs the time-varying target parameter time sequence to the electrical output module. The electrical output module outputs corresponding voltage, current or impedance according to a time sequence, and the communication simulation module constructs a virtual state vector and generates communication data synchronized with electrical output. The environment acquisition module acquires temperature and humidity and binds the temperature and humidity with a time point. And the error fusion module constructs a multi-dimensional error amount and calculates a comprehensive index based on the reference data, the feedback data and the environmental parameters. And the calibration strategy module adjusts a calibration scene or a target parameter track according to the comprehensive index, and outputs a calibration result when a condition is met.

The invention provides an animation production system and method, and the method comprises the steps: recognizing an interactive semantic region, which interacts with a role, in a static background image in an initial animation production material; detecting potential contact points between the roles and the interactive semantic region according to the motion trails of the roles in the role animation sequence and the geometric parameters of the interactive semantic region; when the potential contact point is detected, simulating the deformation influence of the role action on the interactive semantic region based on the preset physical attribute parameter of the interactive semantic region, and generating a corresponding background deformation field; according to the texture features of the interactive semantic region, identifying a synthetic texture of a background region exposed due to deformation, and generating a confidence background sequence synchronously evolved with role actions by a background deformation field and the synthetic texture; and outputting a target animation slice based on the character animation sequence and the confidence background sequence. According to the technical scheme provided by the invention, credible dynamic interaction between the role and the background environment can be realized in the animation production process.

The invention relates to the technical field of model design optimization, in particular to a parameterized simulation model design method and system suitable for a plastic mold. The method comprises the following steps: acquiring parameters of a mold structure, a parting surface and a pouring system through a geometric parameter input module, converting the parameters into an electric signal group and generating a parameter mapping signal; obtaining a mold physical response signal according to the voltage change and forming a visual voltage curve; and performing parameter combination storage through the stable interval and the change rate of the curve, and outputting to a design system in real time to generate a parameterized simulation model. According to the method, the geometric parameters of the plastic mold are electrically signalized, the parameter coupling mapping relation is established, and in combination with precise recognition of the spatial dependency boundary and the parting contact band, intelligent analysis and high-precision parameterized simulation modeling of the mold structure parameters are realized, so that the automation degree and simulation precision of model generation are remarkably improved.

The invention belongs to the technical field of analog simulation, and relates to an AUTOFORM-based hot bath forming process simulation method, which comprises the following steps of S1, establishing a hot forming steel material card; s2, a hot bath forming process route is formulated, wherein the hot bath forming process route comprises a heating section, a hot bath forming section and a cooling section; s3, hot bath forming process parameters are input, wherein the hot bath forming process parameters comprise the environment temperature, the heat transfer coefficient, the friction coefficient and the hot bath time; s4, analogue simulation is conducted on the part, and forming parameters are obtained; and S5, part trial-manufacturing verification is conducted, and forming parameters are adjusted according to part forming results till the parts meet the technical requirements. Based on the AUTOFORM platform occupying 80% of the market share, the hot bath environment is simulated by calibrating parameters such as the heat transfer coefficient and the hot bath friction coefficient during hot bath forming, simulation of the hot bath forming process is achieved, and a basis is provided for large-scale safe application of galvanized hot-formed steel.

The invention provides a method and a system for establishing a jet flow noise prediction model based on a Slowharf number, which are used for realizing high-precision prediction of a jet flow noise spectrum by fusing high-fidelity flow field simulation data, physical dimensionless parameters (such as the Slowharf number) and a deep learning model. The method comprises the steps of parameterized simulation, physical feature extraction, deep neural network training, embedded deployment and the like, and the prediction precision and the calculation efficiency are remarkably improved.

The present invention provides a mechanical learning method of a learning model that outputs a variation parameter in a case where observation data obtained by observing a physical quantity of an actual molding using a molding machine is input, the variation parameter being a variation parameter of a molding condition of the molding machine that reduces a defect degree of a molded product obtained by the actual molding. The mechanical learning method includes the steps of: setting a variation parameter and a fixed parameter to a fluid analysis device, simulating a molding process; acquiring a defect correlation parameter correlated with a defect degree of a molded product obtained by the simulation; calculating the defect degree of the molded product based on the acquired defect correlation parameter; and causing the learning model to perform mechanical learning using the variation parameter set to the fluid analysis device and a reward corresponding to the calculated defect degree.

The invention discloses a parameterized simulation modeling method and equipment suitable for flow field optimization of a thin-walled cylinder, and relates to the technical field of numerical simulation, the method comprises the steps that a grid simulation model comprising a furnace body and a component located in the furnace body is generated, and the furnace body comprises a tuyere; acquiring initial values of geometric parameters of the tuyere; executing a flow field solving operation: solving the flow field distribution of the target model through a solver based on the tuyere geometric parameters and preset simulation time and time step length; after the flow field solving operation is executed, if the flow field distribution of the solved target model does not meet the target requirement, obtaining the adjustment value of the geometric parameters of the tuyere, and returning to execute the flow field solving operation and the subsequent steps until the flow field distribution of the solved target model meets the target requirement; according to the method, the calculation amount and the cost of simulation modeling for flow field optimization of the vacuum high-pressure gas quenching process of the thin-walled cylinder can be reduced, and the optimization efficiency is improved.