15 results about "Global sensitivity analysis" patented technology

This invention relates to the field of geotechnical engineering monitoring and slope stability analysis technology, specifically to a method and system for the layout of slope monitoring points throughout the entire construction and operation cycle. The method acquires multi-source monitoring data and constructs a unified data warehouse through cleaning, format conversion, and spatiotemporal registration. Based on this unified data warehouse, a dynamic risk assessment model coupled with the limit equilibrium method and the finite element method is used to identify target risk areas, and key monitoring variables are identified through global sensitivity analysis. Finally, a multi-objective optimization algorithm is used to solve for the Pareto optimal solution set to determine the location and density of monitoring points. This invention achieves adaptive matching of the monitoring network to changes in geological conditions and dynamic engineering needs throughout the entire cycle, improving the targeting of monitoring and the efficiency of resource allocation.

This paper presents a sparse grid polynomial-based orbital game situation analysis method, belonging to the field of aerospace orbital adversarial and intelligent decision analysis technology. Addressing the problems of lack of systematic representation of multi-source uncertainties, low computational efficiency of traditional quantification methods, and difficulty in achieving multi-objective equilibrium optimization in existing high-orbit multi-to-multi dynamic adversarial scenarios, this paper constructs a POMDP multi-to-multi dynamic game model incorporating uncertainties in high-orbit dynamics and control execution, strategy behavior, and communication delay. A surrogate model is established using sparse grid sampling and polynomial chaotic expansion, and global sensitivity analysis is performed. Based on this, a multi-objective optimization model is constructed, and the Pareto optimal policy set is solved. This achieves game strategy generation that balances performance, robustness, and economy, and is applicable to spacecraft strategy optimization and space situation analysis in high-orbit multi-to-multi dynamic adversarial missions.

The application discloses a kind of solid oxide fuel cell multi-parameter life prediction method, steps are as follows: by SHAP sensitivity analysis, global sensitivity analysis is carried out to solid oxide fuel cell operating parameter, and the key operating parameter that is most significant to the performance of solid oxide fuel cell is screened out;DePatch-CNN life prediction model based on improved PatchTST neural network is constructed, and the improvement includes using differential attention mechanism and inserting one-dimensional convolution front end before block operation;The DePatch-CNN life prediction model constructed by using data set training is used;Using model carries out the multi-parameter life prediction of solid oxide fuel cell, and the preliminary prediction result of model is corrected using key operating parameter as correction factor, and the final voltage prediction value is obtained.The application reduces the model input dimension by sensitivity analysis, enhances the noise resistance of model by the synergistic effect of differential attention mechanism and one-dimensional convolution, effectively improves the accuracy and robustness of solid oxide fuel cell life prediction.

The invention discloses a precision distribution method for reducing dead point repeated positioning errors of a toggle mechanism. The method comprises the following steps: modeling a pin shaft hole fit clearance of each hinge pair as a plane error vector constrained by an upper limit; performing multi-group global random sampling and dynamic simulation based on the model, and extracting a minimum value of a displacement curve of the main sliding block to calculate a dead point error; further, a Sobol global sensitivity analysis method is adopted to quantify the independent contribution and the coupling contribution of each hinge pair error to the variance; and finally, by taking a preset dead point repeated positioning precision allowable bandwidth as a constraint, according to the contribution weight, carrying out backstepping and iterative optimization to obtain an error upper limit configuration vector of each hinge pair differentiation. According to the method, the key error source can be accurately identified, and a quantitative and verifiable distribution scheme is provided for the hinge pair tolerance design while the dead point repeated positioning precision is ensured.

The application discloses a point source emission carbon dioxide dry air column concentration remote sensing inversion error estimation method, system, device, medium and product, relates to the atmospheric gas monitoring technical field, and the method comprises the following steps: determining observation scene parameters affecting carbon dioxide remote sensing observation based on radiation transfer theory; constructing a point source carbon emission observation scene, generating radiance simulation data set by using a radiation transfer model; using the step-by-step Sobol analysis method, the iteration combination of global sensitivity analysis and local sensitivity analysis is used to analyze the influence of observation scene parameters on carbon dioxide radiance, and the parameter sensitivity degree is determined; according to the maximum posterior probability estimation algorithm and the benchmark scene parameters, the column average kernel and the posterior error are calculated, the posterior inversion error of the point source emission carbon dioxide dry air column concentration is calculated, the inversion error estimation of the point source emission carbon dioxide dry air column concentration is completed, and the effective determination of the point source remote sensing inversion influencing factor type and the quantitative estimation of the influence degree are realized.

This invention relates to a method, apparatus, storage medium, and electronic device for obtaining key components of a mixture. The method includes: obtaining each component in a multi-component mixture system in a target region; determining the number of factor levels corresponding to each component; constructing trajectory starting points based on the number of factor levels; determining that the currently constructed trajectory starting points are not repeated based on a previously constructed set of trajectory starting points; determining a preferred set of trajectory starting points based on the uniformity deviation of each constructed set of trajectory starting points; for each preferred trajectory starting point in the preferred set of trajectory starting points, performing a transformation according to the factor order corresponding to that preferred trajectory starting point, with each transformation targeting the factor level of one factor, to obtain a mixture sample; conducting a mixture toxicity test based on the mixture sample to obtain mixture toxicity data; performing a global sensitivity analysis on the mixture toxicity data; and determining key components based on the results of the global sensitivity analysis. This can improve the reliability of mixture toxicity assessment results.

The application provides a regional crop planting management method and device, and applies to the technical field of agricultural production management. The method comprises the following steps: determining a target crop according to a crop growth scene of a target region; performing global sensitivity analysis on a crop growth model based on physiological parameters of the target crop, soil parameters of the target region and management parameters, obtaining sensitive parameters, and performing localized calibration on the crop growth model based on the sensitive parameters; coupling a calibrated crop growth model with a pesticide transport model and a nutrient transport model to construct a comprehensive growth simulation system; combining soil parameters and cost factors of the target region; and traversing and simulating environmental and income changes under different planting strategies through the comprehensive growth simulation system to determine a target crop planting strategy.

The application discloses an engine parameter sensitivity analysis method and related equipment, and relates to the field of rocket engines. By combining qualitative and quantitative analysis, the design parameters that significantly affect the performance deviation of the engine are identified, and a corresponding deviation propagation model is established. First, a performance calculation model based on the interior ballistic equation is constructed, a set of design parameters is determined, and a parameter sample space is generated through statistical distribution. A qualitative global sensitivity analysis method is used for preliminary screening to screen out a set of sensitive parameters that affect the performance index. Then, a quantitative global sensitivity analysis method based on variance decomposition is used to calculate the sensitivity index of each parameter. Finally, the propagation relationship between the parameter deviation and the performance deviation is accurately described through sorting and establishing a deviation model. The application effectively improves the comprehensiveness and accuracy of the sensitivity analysis, and can provide scientific guidance for engine design optimization, performance prediction and deviation management.

The present application is suitable for the technical field of building energy consumption analysis, and provides a building energy consumption dynamic response analysis method and system under meteorological evolution, which comprises the following steps: integrating a multi-source meteorological data set and constructing a building energy consumption reference model; generating sample combinations of uncertain parameters by using a sampling method, running the building energy consumption reference model, obtaining energy consumption output results under different climate scenarios, and statistically analyzing the energy consumption output results; based on the energy consumption output results, using a global sensitivity analysis method to analyze the contribution of each uncertain parameter to the energy consumption, and identifying key driving parameters; based on the sample set generated in the uncertainty analysis step, using a machine learning algorithm to train a building energy consumption prediction model, and outputting building energy consumption prediction results for target meteorological conditions. The present application solves the problem that the traditional method is inaccurate in evaluating under non-stationary climate, and can provide quantitative and robust decision support for climate-adaptive building design and energy-saving strategy formulation.

This invention discloses a method and system for uncertainty analysis of turbine blade tip aero-thermal performance. The method first identifies key structural parameters of the turbine blade tip as uncertainty variables and constructs a multi-precision sample set. Then, a multi-fidelity deep neural network model is constructed using cascaded prediction and correction sub-networks, establishing a mapping correction relationship from low to high precision through phased training. Next, the trained blade tip aero-thermal performance prediction network model is used to analyze the aero-thermal performance distribution, and the influence of each uncertainty variable on performance fluctuations is quantified using the Sobol global sensitivity analysis method. Finally, significant parameters are sorted according to their influence and their optimal value combinations are determined, forming a design scheme that makes the turbine blade tip aero-thermal performance approach the global optimum. This invention effectively overcomes the dimensional limitations of high-dimensional uncertainty analysis, significantly improves computational efficiency and prediction accuracy, and provides reliable technical support for robust turbine blade tip design.

This invention belongs to the field of steel structure engineering quality assessment, and particularly relates to a technology and method for assessing the quality of steel structure assembly based on multi-source information fusion. The method includes: constructing a quality assessment index system and risk level standards for steel box girder assembly; evidence expression and basic probability allocation generation based on cloud models; multi-source evidence fusion and decision-making based on improved D-S evidence theory; risk analysis and defuzzification based on Monte Carlo simulation; identification of sensitive factors based on global sensitivity; and risk visualization. This invention, by integrating cloud models, improved D-S evidence theory, Monte Carlo simulation, and global sensitivity analysis, constructs a systematic assessment index system and risk level standards, achieving accurate quantification, multi-source evidence fusion, level determination, trend analysis, and sensitive factor identification of steel box girder assembly quality risks.

The application discloses a die steel rolling multi-field coupling simulation method, system, equipment and medium, and particularly relates to the technical field of metal plastic forming simulation, and comprises data acquisition, feature analysis, primary processing, secondary processing, coupling simulation and decoupling result output. The application realizes real-time feedback of dynamic recrystallization softening effect to stress solving by constructing two-way coupling conditions of temperature field, stress-strain field and microstructure evolution model; global sensitivity analysis and data preprocessing are introduced to realize input feature dimension reduction, improve solving efficiency and stability; adaptive time step and sequential coupling iteration strategy are adopted to enhance the calculation convergence under complex working conditions; the coupling results are decoupled, a mapping atlas is constructed, and structured data and analysis reports are output. The application realizes high-precision and high-efficiency coupling simulation of multi-physical fields in the die steel rolling process, and can be used for process optimization and quality prediction.

The application discloses a kind of multi-branch closed-loop transmission chain precision distribution methods considering precision-rigidity coupling, comprising: clear multi-branch closed-loop transmission chain end multi-station output index, establish geometric error transmission model;Sobol global sensitivity analysis is used to identify key geometric error term and carry out initial precision distribution;Introduce working load, establish elastic structure equation, analyze the influence of load transmission and elastic deformation on end error;Multi-objective optimization model containing position accuracy, attitude accuracy and structure stiffness cost is constructed, and Pareto solution set is solved and the optimal scheme is selected using TOPSIS method;Output geometric error final allowable value, structure stiffness configuration and manufacturable engineering control index.The application realizes the collaborative design of geometric precision and structure stiffness, and significantly improves the end execution precision of transmission chain and design economy.

This invention provides a digitally driven method for optimizing the protective performance of composite structures, relating to the field of composite material performance optimization. The method includes: multi-scale digital modeling of composite protective structures, construction of high-fidelity simulation models, iterative construction of a performance database based on adaptive sequence sampling, surrogate model construction and global sensitivity analysis, multi-objective global optimization, and optimal solution selection. This method effectively reduces the reliance on empirical data during the design process, reduces the number of penetration tests, lowers testing costs, and shortens the design cycle. It achieves an optimal balance between the dual objectives of "minimizing areal density and maximizing ballistic limits," thereby rapidly designing and optimizing high-performance ballistic composite materials that meet protection specifications.