14 results about "Experimental data analysis" patented technology

The invention discloses a memory alloy part durability experiment data analysis method and system, relates to the technical field of data processing, and aims to solve the problem of poor alloy performance detection effect caused by environmental noise interference and signal scattering of existing memory alloy durability experiment data. A differential signal processing strategy is adopted to carry out noise reduction optimization on multi-dimensional performance response data, characteristic factors representing deformation recovery, residual deformation and energy dissipation are extracted and corrected, and combined logic of the characteristic factors is dynamically judged by combining a recognition rule set based on an application scene so as to determine a specific performance degradation mode. Corresponding residual use cycle prediction models are constructed according to different degradation modes, the credibility range is calculated, finally, a performance optimization strategy is generated through matching, and full-process closed-loop analysis from data preprocessing and feature extraction to degradation mode recognition, service life prediction and strategy generation is achieved. And the performance prediction precision and the reliability evaluation efficiency of the memory alloy part under the complex working condition are remarkably improved.

The invention relates to an autonomous attack and defense method and system, which realize large-scale autonomous attack and defense on an artificial intelligence model based on multi-agent cooperation. The method comprises the following steps: designing an intelligent body check and analyzing technical data by using an attack and defense method, identifying potential attack vectors, defense mechanisms and risk modes from the technical data, and generating an attack and defense blueprint; on the basis of the attack and defense blueprint, software package dependence needed by an experiment is recognized through an experiment deployment agent, a virtualization environment is constructed, and loading and configuration of a model and a data set are completed; and executing an attack and defense experiment in a deployed experiment environment by using an experiment debugging agent, collecting experiment data, analyzing model performance and generating an effect evaluation and feedback report, thereby realizing self-adaptive optimization and sustainable evolution of the system. Compared with the prior art, the method has the advantages that efficient autonomy and knowledge driving of security attack and defense are realized, and security risk mining, alignment evaluation and defense verification can be performed on the model and the system without manual intervention.

The invention discloses an experimental data analysis method, device and equipment and a storage medium, and relates to the field of data processing, and the method comprises the steps: carrying out the demand analysis of an experimental data file and a user analysis demand based on a large language model, and determining an experimental task specification which at least comprises an experimental type; according to the experiment type and the experiment data file, generating a standardized data table and a data portrait; determining an experimental data analysis plan according to the experimental task specifications and the data portraits; executing an experimental data analysis plan, and performing data processing on the standardized data table to obtain cartographic data and a statistical analysis result; based on a decoupling mechanism of drawing code generation and execution, generating experimental chart data according to the drawing data; and generating an experimental data analysis report according to the experimental chart data and the statistical analysis result. According to the scheme, end-to-end experimental data analysis from data input to report generation is realized, and the accuracy of experimental data analysis is improved.

The invention discloses a buckling failure behavior prediction method for a laser welding stiffened plate. The method comprises the following steps: S1, preparing a compression experiment sample; s2, carrying out a compression experiment; s3, analyzing compression experiment data; s4, finite element analysis; s5, establishing a machine learning model and training; according to the method, the compression buckling failure of the laser welding stiffened plate is expanded, an experimental method, numerical simulation, machine learning and software development are combined, a sample made of AlLiS4 and 2099 aluminum-lithium alloy is adopted, structural design is carried out to ensure buckling failure, a compression clamp is used for applying a load in an experiment, strain evolution is monitored through a strain gauge and a fringe projection technology, and the deformation of the laser welding stiffened plate is detected. According to data analysis, Abaqus software is adopted to establish a finite element model, a bilinear isotropic strengthening model is applied to predict a buckling load, post-buckling analysis is carried out, multiple technologies are integrated, an innovative solution is provided for design and analysis of the stiffened plate, and wide application potential is shown.

This invention relates to the field of space application technology, and discloses a method, system, and related equipment for processing scientific experimental data on space payload materials. The method includes: collecting load condition data, experimental image data, and material temperature data from scientific experiments on space payload materials, and performing time-series processing; determining the corresponding laser shutdown time points of the scientific experimental payload in the load condition data, experimental image data, and material temperature data; aligning the load condition data, experimental image data, and material temperature data and normalizing the acquisition frequency based on these time points; and performing joint analysis and processing on the various types of data after data alignment and acquisition frequency normalization. This invention, by determining the laser shutdown time points of the scientific experimental payloads for various types of data and aligning and normalizing the acquisition frequency of the data according to these time points, achieves precise correlation among various types of data, greatly improving the efficiency and accuracy of scientific experimental data analysis.

The application discloses an experimental data analysis method and device, equipment and a storage medium, and relates to the field of data processing.The method comprises the following steps: based on a large language model, performing demand analysis on an experimental data file and a user analysis demand, determining an experimental task specification, and determining that the experimental task specification at least comprises an experimental type; generating a standardized data table and a data portrait according to the experimental type and the experimental data file; determining an experimental data analysis plan according to the experimental task specification and the data portrait; executing the experimental data analysis plan, performing data processing on the standardized data table, obtaining mapping data and statistical analysis results; based on a decoupling mechanism of mapping code generation and execution, generating experimental chart data according to the mapping data; and generating an experimental data analysis report according to the experimental chart data and the statistical analysis results.The application realizes end-to-end experimental data analysis from data input to report generation, and improves the accuracy of experimental data analysis.

The utility model discloses an organoid culture pore plate convenient to count and position, which belongs to the technical field of cell culture and comprises a culture pore plate, a plurality of pore grooves are arranged on the culture pore plate, scale marks are arranged at the bottoms of the pore grooves, and the scale marks comprise a plurality of transverse lines and a plurality of longitudinal lines which are uniformly distributed outwards by taking the circle centers of the pore grooves as centers. According to the utility model, orientation coordinates in the pore plate are arranged, so that the dynamic growth process of the organoid can be conveniently positioned and observed, regular counting of each organoid in each pore is facilitated, the budding number is calculated, and the accuracy of statistical analysis is improved; by setting the grids, the organoid can be located in a grid with a fixed area, so that the area can be conveniently calculated without being limited by different shooting machines or software, the efficiency of experimental data analysis is improved, and calculation or statistical errors are reduced; the same type of organs are continuously observed through scale line positioning, the efficiency of organoid culture experiment operation and data analysis and the data reliability are improved, and the device has very high practicability and wide application prospects.

The application discloses an intelligent simulation experiment method and device, and the method comprises the following steps: obtaining task demand information; constructing an intelligent simulation experiment model; and processing the task demand information by using the intelligent simulation experiment model to obtain optimized experiment scheme data. The intelligent simulation experiment model is constructed by a data input intelligent agent, a simulation resource access intelligent agent, a basic scenario generation intelligent agent, an experiment scenario generation intelligent agent, an experiment operation control intelligent agent, an experiment data analysis intelligent agent and an experiment scheme recommendation intelligent agent. The software application in the form of the intelligent agent is used to replace the traditional combat simulation experiment software, and the combat simulation experiment task AI intelligent agent is used to replace manual operation, so that the intelligent combat simulation experiment is realized, and the problems of high personnel requirement, great energy consumption and few innovative conclusions in the traditional combat simulation experiment are solved.

The application discloses a method for predicting the mechanical properties of martensitic heat-resistant steel by using EBSD detection data analysis, adopts area weighted average value to obtain corresponding EBSD data, and overcomes the geometric distortion problem of a traditional equivalent circle diameter method when processing non-equiaxed and slender lath martensite; the concept of martensite lath fitting length-diameter ratio is innovatively proposed, and through a large amount of experimental data analysis, the correlation between the concept and the mechanical properties of the martensitic heat-resistant steel is found; through a large amount of data, a prediction formula between the martensite lath fitting length-diameter ratio and the mechanical properties can be established, so that the mechanical properties of the final product can be accurately predicted by measuring the EBSD data of the sample in the subsequent production process.

The utility model discloses a laboratory data analysis equipment, including electrophoresis appearance main part, set up on electrophoresis appearance main part's plug hole, digital display screen, operating button and electrophoresis tank, electrophoresis appearance main part is provided with the storage mechanism, the storage mechanism includes two wire box that sets from top to bottom, the inside of wire box is provided with transmission wire, coil spring and winding rod. Through setting wire box, coil spring and winding rod, make transmission wire have the tendency of winding in the box after elongation, the storage of transmission wire is convenient, through setting U -shaped board, insert piece and limiting screw, make the storage mechanism and electrophoresis appearance main part quick installation and disassembly convenient, through setting apron, it is convenient to quick maintenance replacement to transmission wire, thereby the experimental data analysis equipment has the effect of the wire of electrophoresis experiment and stores, avoids the mutual entanglement of wire, further reaches the purpose of improving electrophoresis experiment efficiency.

The invention discloses a conditional diffusion model image semantic editing method based on dynamic mask prediction, and belongs to the technical field of computer vision. Comprising the following steps: step 1, early-stage preparation: constructing a model and initializing parameters; step 2, training a mask prediction module; 3, carrying out initialization and editing fusion operation on an image editing framework; (3a) forward noise addition and background information extraction; (3b) de-noising generation and image fusion of editing branches; step 4, feature extraction and dynamic mask prediction; (4a) feature extraction; (4b) vector construction and mask generation; 5, performing loop iteration to obtain a final editing result; according to the method, a reconstruction branch is additionally arranged on the basis of an original diffusion model, a dynamic mask prediction module is designed, and accurate definition and dynamic adaptation of an editing area are achieved. Through experimental data analysis, the method provided by the invention can effectively reduce FID and Local FID values, improve CLIP Score, and obtain a better editing effect.

The invention discloses a dynamic experimental analysis method and device based on calibration type flow reaction calorimetry. The method specifically comprises the following steps: firstly, carrying out a flow reaction calorimetric experiment to obtain experimental data, then, carrying out calorimetric analysis to obtain a thermal conversion rate of each section, calculating reaction rate constants of a reaction process under different temperature conditions by utilizing a nonlinear fitting mode, and further, calculating kinetic parameters of the reaction process by utilizing a linear fitting mode. According to the method, additional component detection equipment is not needed, pre-exponential factors and activation energy in the reaction process are obtained through analysis of experimental data under different temperature conditions, dynamic analysis can be carried out while dynamic feeding chemical reaction calorimetric analysis is carried out under the continuous flow process condition, the experimental efficiency is high, and the process is safe.

The invention relates to a chaos engineering experiment data analysis method and device based on an AI large model, and relates to the technical field of data processing, and the method comprises the steps: obtaining target fault data and target system operation data of a chaos engineering platform; the target system operation data is operation data of the application system after the chaotic engineering platform injects the target fault data into the application system; calling an API interface of the large language model according to the target fault data and the target system operation data to obtain a target data analysis result returned by the large language model; providing the target data analysis result to a chaos engineering platform; wherein the target data analysis result is determined in combination with the large language model, analysis and processing by analysts can be avoided, and the data analysis cost is reduced; and the data analysis efficiency can be improved based on the large parameter quantity and accuracy of the large language model.

This invention discloses a method and system for optimizing hot melt adhesive preparation process parameters based on processing feedback, relating to the field of hot melt adhesive preparation technology. The method includes: classifying hot melt adhesive formulations and setting corresponding formulation labels; conducting single-factor group preparation experiments for different formulation labels; plotting process sequence curves and generating preparation trajectory curves based on single-factor experimental data; extracting performance parameters of each process factor and identifying response characteristics; dividing the entire preparation process into multiple stages; constructing a collaborative optimization prediction model for process parameters based on a neural network, using formulation labels, process parameters, performance parameters, and response characteristics as inputs, and outputting preparation time reduction rate and performance achievement rate; and testing the model's predictive reliability for new samples. The advantages of this invention are: by combining formulation classification, single-factor experimental data analysis, and neural network modeling, intelligent optimization of hot melt adhesive preparation process parameters is achieved.