2317 results about "Performance prediction" patented technology

The invention relates to the technical field of concrete detection, in particular to a dynamic detection and analysis method and system for early performance of ultra-high performance concrete. According to the technical scheme, the method comprises the steps of establishing a raw material characteristic spectrum library based on multi-physics field coupling, constructing a dynamic detection array based on a time sequence, developing a multi-modal data fusion algorithm, establishing a nonlinear coupling effect decomposition model and implementing dynamic parameter inversion. According to the method, a technical chain of four-dimensional monitoring network-multi-physics field coupling analysis-cross-scale prediction-adaptive regulation and control is constructed, so that dynamic accurate detection and intelligent regulation and control of the early performance of the ultra-high performance concrete are realized, and the technical bottlenecks of a traditional method in the aspects of parameter coupling, cross-scale association and real-time regulation and control are solved; the early performance prediction precision and the maintenance efficiency are remarkably improved, the cracking risk is reduced, and reliable technical support is provided for application of the ultra-high performance concrete in major engineering.

A transformer-based student performance prediction and reasoning intervention is disclosed. The system comprises a data repository coupled to a transformer-based prediction module that processes student data through multi-head attention mechanisms to generate performance predictions and identify potential learning shortfalls. A reasoning-enhanced large language model algorithmically generates personalized corrective action plans by applying structured decomposition of learning challenges, multi-step reasoning, and hypothesis testing. An algorithmic prompt formulation system optimizes inputs using field-specific, level-specific, and shortfall-specific templates. The system implements a workflow including shortfall detection against educational thresholds, causal factor analysis, intervention generation, and adaptive refinement based on outcomes. This approach enables early identification of academic challenges and timely implementation of personalized interventions to improve student learning outcomes.

The invention provides an intelligent analysis and execution method and system for database cascading operation, and relates to the technical field of database management.The intelligent analysis and execution method comprises the steps that a target table and an association table are extracted through a recursive query algorithm, field mapping analysis is carried out to generate a dependency weight matrix, cyclic dependency is recognized based on the matrix and decomposed into directed acyclic subgraphs, and the directed acyclic subgraphs are analyzed; node depth values and breadth values are calculated for topological sorting to generate an execution sequence; performing node classification on the execution sequence, organizing the execution sequence into a batch processing task group, calculating an optimal execution path through an adaptive path optimization algorithm, generating a distributed transaction control instruction, and allocating the task group to an execution thread pool for execution by using a consistent Hash algorithm; monitoring an execution process, dynamically adjusting thread pool resources according to a performance prediction value, executing cascade rollback when an exception occurs, and reconstructing a transaction execution path; according to the method, the execution efficiency and stability of the database cascade operation can be effectively improved, and the system resource consumption is reduced.

The invention discloses a valve performance simulation and optimization method and system based on mixed deep learning, and the method comprises the steps: S1, carrying out the finite element simulation analysis of a valve, and solving through the thermosetting coupling of a multi-physics field, stress, deformation and temperature distribution data of the valve under different working conditions and maximum thermal stress and maximum thermal deformation displacement of key parts of the valve are obtained; s2, after data processing, taking stress, deformation and temperature distribution data of the valve under different working conditions as input parameters, taking maximum thermal stress and maximum thermal deformation displacement as output parameters, training a neural network model, and establishing a thermal-structural performance prediction model; s3, on the basis of a prediction result, a reference point-based non-dominated genetic algorithm is adopted, a reference point generation and self-adaptive agent model is combined, and multi-objective optimization is carried out to output an optimal parameter combination so as to adjust valve design; according to the invention, deep learning and multi-objective optimization are carried out on the finite element simulation result by using the neural network, and the valve performance prediction precision and optimization efficiency are improved.

The invention relates to the technical field of electrical design and intelligent optimization calculation, and discloses a capacitor structure design optimization method and system and a storage medium. The method comprises the following steps: carrying out modeling processing on capacitor structure parameters, and dividing a parameter space to obtain an initial model set; obtaining a multi-physical field simulation model according to the geometric model and the material attributes; obtaining performance indexes such as capacitance value, heat distribution and stress based on the simulation model; setting a target function and constraint conditions, and operating an optimization algorithm to obtain an optimal solution set; performing feedback control processing on an optimization result, and analyzing a convergence path to obtain a structure parameter; according to the invention, the execution efficiency of capacitor structure design optimization is improved, and the consistency and stability of the optimization process and the performance prediction result are improved.

The invention provides a comprehensive financial IT operation and maintenance management system and method based on artificial intelligence, and the system comprises an intelligent monitoring and predictive maintenance module which is used for achieving the real-time monitoring, abnormality recognition and performance prediction of an IT system through collection of multi-dimensional time series data, and when an abnormal condition is monitored or a prediction result does not meet a preset requirement, the intelligent monitoring and predictive maintenance module carries out the operation and maintenance of the IT system; if so, triggering intelligent early warning; the AI-driven fault diagnosis and self-healing module is used for carrying out intelligent fault analysis and repair in combination with IT system logs and historical fault data based on the detected abnormal condition; the self-adaptive resource management module is used for predicting a future load trend and a future performance trend, and performing self-adaptive resource management based on the predicted future load trend and the future performance trend in combination with cost and energy consumption factors; and the security operation and maintenance automation module is used for monitoring security threats of the IT system in real time and automatically executing security strategies.

The invention relates to the technical field of building simulation, and discloses a BIM-based special-shaped hyperboloid building site model generation method and system, and the method comprises the steps: collecting the multi-source heterogeneous point cloud data of a special-shaped hyperboloid building site to construct a three-dimensional feature tensor model, and recognizing the key curvature mutation point of the tensor model to construct a space reference topology network; based on the digital skeleton model, geometric contour features are analyzed, hyperboloid fitting is carried out, and an initial BIM model is obtained; performing multi-physics field coupling simulation on the initial model to obtain a target BIM model; extracting curvature distribution and structural stress parameters of the target BIM model to obtain basic model parameters, and constructing a building performance prediction model in combination with historical performance reference data; and inquiring a bearing capacity standard and a deformation threshold value, and analyzing the optimal curved surface parameter combination and a prestress applying strategy to obtain optimized simulation parameters, so as to simulate the special-shaped hyperboloid building by utilizing target model simulation. According to the method, the simulation precision of the special-shaped hyperboloid building can be improved.

The invention relates to the technical field of process optimization, and discloses a hot work die steel heat treatment process optimization method and system.The method comprises the steps that hot work die steel samples are collected under multiple sets of different process conditions, and a performance basic data set is obtained; constructing a multi-target coordination optimization model based on the performance basic data set; performing phase change detection on the hot work die steel sample to obtain phase change monitoring data; performing prediction in combination with the phase change monitoring data to obtain a performance prediction result; and solving an optimal process parameter combination based on the performance prediction result and the multi-target coordinated optimization model, and realizing simultaneous optimization and coordinated balance of a plurality of performance indexes in the heat treatment process of the hot work die steel by making full use of associated information among different performance indexes.

The invention relates to a hybrid architecture mechanical property prediction method and system for an additive manufacturing lattice structure, and the method comprises the steps: converting a lattice structure node three-dimensional coordinate into a normalized distance from a lattice center, introducing a self-attention mechanism, and constructing a feature prediction model in combination with residual connection and layer normalization; constructing a generative adversarial network comprising a generator and a discriminator; a strategy network based on prediction error rewards is constructed, and a depth deterministic strategy gradient algorithm is adopted for optimization; jointly training the models, and constructing a performance prediction feedback model; finally, the node coordinates of the target structure serve as input, and mechanical property prediction output is achieved. According to the method, the generative adversarial neural network, reinforcement learning and an attention mechanism are mutually combined to generate a hybrid architecture, the problems of gradient disappearance and mode collapse of a traditional GAN can be effectively relieved, the model can accurately recognize implicit association between node positions and mechanical properties, and then the mechanical properties of target lattice structure parameters are rapidly predicted.

The invention discloses a valve vibration impact test method and system based on digital twinning, and belongs to the technical field of industrial equipment test and evaluation. The method comprises the following steps: firstly, deploying a multi-source sensor network, constructing a physical entity layer of a valve vibration impact test, collecting and preprocessing various physical monitoring data, and constructing a digital twinborn body layer; secondly, real-time data assimilation with a physical test bed is realized through parameter calibration and updating of a digital twinborn model based on various physical monitoring data and the digital twinborn layer; and finally, executing a virtual vibration impact test, realizing prediction and evaluation of global response and key part information of the physical entity according to a test result, and finally realizing interaction and performance prediction. And the valve vibration impact test is realized through the valve vibration impact test system. According to the method, the fault diagnosis capability, the early damage identification capability and the valve state prediction and evaluation capability can be remarkably improved, and the defect of insufficient field monitoring information is overcome.

The invention is suitable for the technical field of performance detection, and provides a method for detecting the ablation resistance of a burning-resistant nozzle and an arc contact, and the method comprises the steps: collecting a temperature field matrix, an arc image sequence, an electrical parameter time sequence and a surface displacement curve of an ablation region, and generating a multi-modal data set; based on the data set, extracting spatial features of the ablation core region through a cavity convolution layer and a channel attention mechanism; inputting a preset bidirectional LSTM network, fusing thermal stress data of the COMSOL simulation model to drive gating weight update, and extracting time sequence evolution characteristics; generating a space-time fusion feature vector containing the temperature field gradient, the arc form change rate and the surface roughness evolution feature; inputting the space-time fusion feature vector, the thermal stress distribution data and the material phase change threshold parameter into a constructed ablation prediction model for training; the ablation prediction model after training is completed is optimized; and a performance prediction result is generated in real time by using the optimized ablation prediction model, so that the detection accuracy is effectively improved.

The invention provides an AI-enhanced large mine water-doped concrete material performance prediction and mix proportion design integrated system, aiming at the key problem of large mine water-doped concrete material temperature control, the importance of different influence factors such as raw material-structure-environment and the like is comprehensively and quantitatively evaluated; aiming at a complex raw material system of the large mine water-doped industrial concrete, main features of a performance prediction model such as feature dimension reduction and weight calculation are optimized, and meanwhile, a machine learning algorithm combined with a self-attention mechanism is constructed so as to reduce interference of features weakly related to target performance, and the performance regulation and control capability of the large mine water-doped industrial concrete is enhanced. Furthermore, a multi / super multi-target optimization algorithm containing a constraint system is developed and integrated aiming at targets such as material key performance, carbon emission and cost in the preparation process, and a comprehensive and flexible data-driven solution is provided for temperature control, influence factor evaluation, performance regulation and control and proportion optimization of the large-mine water-doped concrete material. Therefore, the overall decision-making and design process is optimized.

The invention discloses an insulating material performance evaluation and formula optimization method and device and a medium, and the method comprises the steps: extracting multi-dimensional data features according to the multi-physical field data of an insulating material, carrying out the feature fusion, and obtaining multi-modal fusion features; constructing a material performance prediction model based on the multi-modal fusion features, and responding to real-time multi-physical field data to obtain a residual life prediction result of the insulating material; determining a key factor influencing the life based on the residual life prediction result, and mapping data characteristics of the key factor to an insulation material failure mechanism to determine a failure factor weight; based on the failure factor weight, establishing a formula optimization objective function by adopting a multi-objective reinforcement learning algorithm; solving the formula optimization objective function based on constraint conditions to obtain a material formula optimization result; the accuracy of the residual life prediction result of the insulating material is improved, and meanwhile, the reliability and pertinence of the formula optimization result of the insulating material are considered.

The invention discloses an air flotation vacuum pump performance prediction system based on a specific calculation model, and relates to the technical field of mechanical engineering. Comprising a multi-parameter fluctuation sensing module, a weak feature sensitive extraction module, an abnormal coupling atlas construction module, an instability precursor identification module, a prediction model regulation and control module and a model adaptive optimization module, and carrying out fusion in a preset time window, constructing a multi-parameter synchronous fluctuation sensing model, and outputting a synchronous anomaly index. By introducing multi-parameter synchronous fluctuation perception, weak feature extraction and an abnormal coupling map, accurate identification and early warning of the early instability trend of the air floating vacuum pump are realized, and an intelligent system with real-time perception and closed-loop prediction capabilities is constructed by combining model regulation and control and a self-adaptive optimization mechanism, so that the operation safety and stability are improved.

The invention belongs to the technical field of field effect transistor performance prediction, and particularly relates to a reliability prediction method, device and system for a field effect transistor, and the method comprises the steps: determining the correlation between the time sequence data of environment parameters, electrical parameters and performance parameters in historical operation records and the change of the performance parameters through a correlation analysis method when the time sequence data changes; key performance parameters are screened out, then a virtual operation model of the field effect transistor is constructed before life prediction is conducted through the screened key performance parameters, and historical data are used for conducting validity verification on the virtual operation model; finally, dynamic changes of environment parameters and electrical parameters are simulated through historical data when the successfully verified virtual operation model is utilized, the degradation trend of key performance parameters is recorded, residual life prediction is carried out based on the degradation trend, prediction efficiency is improved, life prediction can more accurately reflect actual working conditions, and the service life prediction efficiency is improved. And the accuracy of residual life prediction is improved.

The invention provides a data correction method and system combined with lattice structure additive manufacturing process characteristics, and the method comprises the steps: firstly employing a Newton iteration method, taking the radius of a pillar as a variable, optimizing the relative density of an iteration target and generating a lattice structure geometric model under the condition that a process constraint condition is satisfied, and extracting actual geometric parameters after forming through CT scanning, the elastic modulus is corrected through the pillar diameter deviation and the defect volume fraction, a compression failure mechanism is combined, a density-related failure criterion is introduced, a nonlinear relation with the relative density is constructed through specific energy absorption data, material plasticity parameters are inversely optimized, and a defect coupling evaluation model is constructed according to the surface powder sticking rate and the pillar diameter deviation. And establishing a Gaussian mixture model based on a stress-strain curve to screen out abnormal data, and finally fusing the parameters to construct a data correction model. The dot matrix structure design precision can be improved, and then a high-quality data basis is provided for performance prediction-structure design two-way feedback.

The invention relates to a multi-agent-based material performance prediction and synthesis system, and the system comprises a multi-agent data enhancement module which is configured to be used for firstly disassembling a complex problem into a plurality of subtasks, and then constructing a fine tuning data set comprising Sub-CoQ question and answer pairs by starting multi-source parallel retrieval; the multi-expert debate module is configured to be used for simulating decision conflicts of different roles in material engineering and generating a direct preference optimization DPO data set through debate; the training and verification module is configured to be used for training and verifying a large model MatMind in the field of materials by utilizing supervised fine tuning SFT and reinforcement learning RLHF based on the fine tuning data set and the DPO data set; and the material performance prediction and synthesis module is configured to be used for realizing intelligent recommendation of a material performance prediction and synthesis process by importing input parameters into the large model MatMind.

The invention discloses a blade repairing method and system based on a laser cladding simulation model, and relates to the technical field of blade repairing, the laser cladding simulation model is established, high-speed laser cladding process parameters are set as model variables to be input into the laser cladding simulation model, and the laser cladding process under different parameter combinations is simulated. The method comprises the following steps: preparing a cladding layer, acquiring a simulation result of the cladding layer, preparing cladding samples under different technological parameter combinations, performing characteristic analysis, recording related analysis data, and establishing a performance prediction model of a corresponding mapping relationship between the technological parameter combinations and blade coating performance based on the related analysis data so as to predict the coating performance of the blade under different technological parameter combinations. Combining specific service environment conditions of the blade, setting performance constraint conditions, optimizing a process parameter combination, determining an optimal preparation process parameter combination, performing laser cladding repair on the actually damaged blade, monitoring the laser cladding process in real time, determining whether to intervene or not, and performing quality detection after the repair is completed.

The invention discloses an environment adaptability self-optimization method and system of an intelligent sensing robot, and relates to the technical field of robot control, and the method comprises the steps: obtaining a historical environment data set; acquiring a historical performance data set of the target robot; obtaining an optimal matching relationship; constructing a feature data set; constructing an environment-performance prediction model based on the feature data set; generating predicted performance data of the target robot; and optimizing and adjusting a strategy according to the predicted performance data. The technical problems that in the prior art, an intelligent sensing robot cannot dynamically and accurately predict self performance according to different complex and changeable environments, so that the environment adaptability of the intelligent sensing robot is insufficient, and diversified scene requirements are difficult to meet are solved, the performance of the intelligent sensing robot in the different environments is predicted, and the intelligent sensing robot is more intelligent. And optimization and adjustment can be carried out as required based on a prediction result, so that the method can better adapt to the environment, and the performance is improved.

The invention discloses a common concrete and ultra-high performance concrete performance prediction and mix proportion multi-objective optimization design method and system, and the method comprises the steps: obtaining a comprehensive data set which covers the mix proportion, performance data and cost data of common and ultra-high performance concrete; correlation analysis is carried out, and influence factors strongly correlated with the target variable are determined; constructing a concrete performance prediction model by using a machine learning algorithm, training and testing, then carrying out interpretability analysis, and selecting an optimal concrete performance prediction model; constructing a mix proportion optimization design objective function and an ultra-high performance concrete cost function; establishing a multi-objective optimization design mathematical model in combination with constraint conditions, a cost function and an optimization objective function; and solving by adopting a multi-objective evolutionary algorithm, generating an approximate Pareto frontier solution set, and determining an optimal mix proportion optimization design scheme. According to the method, the concrete performance can be predicted, the mix proportion with the optimal comprehensive performance and cost effectiveness is obtained, and multi-objective optimization design of the concrete mix proportion is achieved.

The invention discloses a deep learning-based safety state evaluation method for a concrete-filled steel tube node. The method comprises the following steps of: 1, constructing a comprehensive safety evaluation index of the concrete-filled steel tube node; 2, collecting dynamic performance data of the nodes, and performing denoising, normalization and feature extraction processing to form multi-dimensional normalized time sequence data; 3, calculating the weight of each index, and optimizing weight distribution; 4, inputting the multi-dimensional normalized time sequence data into a deep learning model, carrying out model training, verification and optimization in combination with optimized weight distribution, and outputting a preliminary comprehensive safety score and a long-term performance prediction result of the node; and 5, further comprehensively evaluating the safety state of the node, determining a final comprehensive safety score, and grading to formulate corresponding grading maintenance suggestions. And 6, dynamically predicting the future performance change trend of the node, and quantifying the fatigue life, the crack growth rate and the material degradation condition of the node. The method is suitable for dynamic safety evaluation and long-term performance prediction of the concrete-filled steel tube node.

Embodiments of the present invention relate to the fields of large model techniques and big data. Disclosed are a data scheduling method, a system, and an electronic device. The method comprises: in a performance feature set of a performance prediction model, determining a target performance feature matched with a scheduling strategy of a model to be trained, wherein the performance prediction model is matched with the model type of the model to be trained, the model type is used for representing a data architecture of the model to be trained, and the scheduling strategy is used for at least representing a rule required to be satisfied during training of the model to be trained; and identifying from among different training strategies a target training strategy corresponding to the target performance feature, and identifying from among different training resources a target training resource corresponding to the target performance feature; and configuring the identified target training strategy to the model to be trained, and allocating the identified target training resource to the model to be trained.

The invention provides an airfoil profile aerodynamic performance prediction method based on a large model retrieval enhancement generation framework, and the method comprises the steps: building a weakly-coupled database through fusing the geometric and aerodynamic characteristics of an airfoil profile, further building a strongly-coupled airfoil profile data distributed retrieval model, building the mutual cooperation of two stages of rough arrangement and fine arrangement, and achieving the prediction of the aerodynamic performance of the airfoil profile. And various characteristics of the airfoil profile are comprehensively considered, so that data can be retrieved more accurately. Meanwhile, fine-grained airfoil aerodynamic knowledge is generated in combination with a large-parameter language model, and the knowledge is optimized through knowledge distillation and fine tuning technologies. According to the method, explainable airfoil aerodynamic design priori knowledge can be excavated, rapid intelligent design of airfoils is powerfully supported, and the method has important significance in promoting the airfoil aerodynamic design to develop towards the intelligent and efficient direction.

The invention discloses a multi-source data fusion foam concrete construction process monitoring method and system, and relates to the technical field of concrete construction, the method comprises the following steps: obtaining a standard response data set of foam concrete, including raw material ratio parameters and performance index parameters which are stored in an associated manner; constructing and training a hybrid hierarchical performance prediction model, wherein the model comprises a shared prediction layer based on integrated learning and a plurality of independent output layers connected with the shared prediction layer; in combination with a multi-objective optimization algorithm, by taking compressive strength and cost optimization as an objective, constructing an evaluation function for global optimization, and obtaining a target mix proportion scheme; acquiring real-time process data associated with the target mix proportion scheme, and inputting the real-time process data into the mixing layering performance prediction model to obtain a prediction result; and comparing with a preset design target, and generating an adjustment instruction for adjusting the matching parameters of the subsequent stirring batches. The problem that in the prior art, the performance fluctuation of foam concrete of different batches is large due to the fact that the construction condition cannot be dynamically adjusted in real time is solved.

The invention provides a composite material performance prediction and process optimization method based on a neural network, and the method comprises the steps: firstly collecting multi-source data in the preparation and test process of a composite material, carrying out the preprocessing of the data, screening key feature variables as input variables, constructing a feedforward artificial neural network model, and predicting and outputting the performance indexes of the composite material. And training the model, performing iterative optimization on model parameters, and optimizing composite material process parameters by using the optimized model based on a reverse optimization strategy of a genetic algorithm to obtain an optimal process parameter combination. The invention provides a scientific, efficient and reliable tool for design and optimization of composite materials, and particularly has wide application prospects in high-requirement industries such as aerospace and the like.

The invention provides a high polymer material performance prediction method based on deep learning, and relates to the technical field of material performance prediction.The method comprises the specific steps that variable parameters influencing high polymer material performance are determined, multiple sets of samples are prepared and subjected to a tensile test and an accelerated aging test, and initial and final tensile performance parameters are obtained; and constructing and training an initial performance prediction model and an aging performance prediction model based on a deep learning network model, inputting related parameters into the models to obtain performance parameters, generating an environment severe coefficient through data processing, calculating a performance evaluation index in combination with a tensile performance parameter variation, and comparing the performance evaluation index with a preset threshold value to evaluate the material performance. According to the method, the performance change of the aged material can be accurately predicted, the performance evolution can be reflected, the prediction result is closer to reality, the material performance can be comprehensively evaluated, and an effective reference and a scientific basis are provided for long-term use, selection and application of the material.

The invention provides a computer hardware performance evaluation method and system based on an analogue simulation model, and relates to the technical field of computer performance evaluation. According to the method, hardware component models such as a processor, a memory and I / O are constructed by analyzing a system structure description file, semantic analysis is performed on an instruction stream or a parallel computing graph of a to-be-evaluated application, and a task load model is formed. The model is input into an event-driven scheduling module, simulation is executed on the hardware component model, task-resource mapping is generated, and task time delay, communication traffic and power consumption are recorded. And calculating average response time delay, data bandwidth and unit energy consumption according to the simulation data, and outputting an evaluation report compared with the performance baseline. According to the method, real load-oriented multi-dimensional performance prediction can be realized, and the method can be used for architecture optimization and scheme selection.

The invention provides an aircraft aerodynamic configuration design method and device based on simulation, and relates to the technical field of aircraft aerodynamic configuration design, and the method comprises the steps: obtaining aerodynamic characteristic parameters of aircrafts of different airfoils to construct a multi-source aerodynamic database; a neural network model is established based on the database, airfoil geometric parameters serve as input, the lift coefficient, the resistance coefficient and the lift-drag ratio serve as labels for training, and an aerodynamic performance prediction model is obtained; candidate combinations are generated in an airfoil geometric parameter design space, aerodynamic characteristics of the candidate combinations are predicted through a model, a fitness function is optimized by utilizing dynamic weights, and the combination with the highest fitness is selected as an optimized airfoil parameter by adopting a genetic algorithm; and performing simulation verification on the optimization parameters, extracting a simulation result, comparing the simulation result with the output of the prediction model, judging that the optimization combination is effective when an error meets a preset requirement, and completing aerodynamic configuration design accordingly. The method improves the prediction accuracy and design efficiency of aerodynamic performance.

The invention provides a load balancing method and device based on dynamic weight, equipment and a storage medium, and relates to the technical field of load balancing, and the method comprises the steps: obtaining historical node information of each service node after receiving a request of a client; based on the historical node information, calling a node information prediction model to calculate a node information prediction value at the next moment; calling a weight calculation model to calculate a node information weight based on the node information predicted value; performing normalization processing on the node information weight of each service node to obtain a current node information weight; performing weighted summation processing on the node information predicted value by adopting the current node information weight to obtain a comprehensive performance predicted value of each service node; and selecting the service node with the highest comprehensive performance prediction value as a target service node for routing. According to the embodiment of the invention, the weight of each node information is dynamically adjusted based on the node information predicted value, a more adaptive service node can be selected, and the resource utilization rate and stability of the system are improved.

The invention provides a training method and device for a semiconductor device performance prediction model and related equipment, and the method comprises the steps: determining key parameters and performance parameters of a semiconductor device, the key parameters comprise key size parameters and / or process parameters, and the performance parameters comprise electrical characteristic parameters of the semiconductor device; based on the multiple groups of key parameter values and the corresponding performance parameter values, establishing a sample set; the key parameter values are disturbed, and the correlation degree of the key parameters and the performance parameters is determined; wherein after the key parameter value is disturbed, the change condition of the corresponding performance parameter value is used for representing the correlation degree of the key parameter and the performance parameter; initializing the weight of a preset neural network model based on the performance prediction target and the correlation degree of the key parameter and the performance parameter; and inputting at least part of samples in the sample set into a weight-initialized preset neural network model for training to obtain a semiconductor device performance prediction model.