62 results about "Experimental model" patented technology

The invention belongs to the technical field of composite material design, and discloses a polyimide-based composite material design method and system based on experiment-machine learning collaborative optimization. The method comprises the following steps: S1, experimental database construction and machine learning performance modeling; S2, machine learning model construction and training; S3, model evaluation and performance index output; and S4, intelligent design of the polyimide-based composite material. The multi-objective performance collaborative optimization design and preparation of the polyimide-based ternary carbon heterostructure composite material are carried out by taking experimental data as a main material and machine learning as an auxiliary material. According to the method, an experiment-model-optimization closed-loop iterative design system is constructed based on an experiment-machine learning collaborative optimization method, the method has the advantages of high prediction precision, high optimization efficiency, good multi-target adaptability and the like, the development efficiency of the polyimide-based composite material is remarkably improved, and the development cost of the polyimide-based composite material is reduced. The method is suitable for intelligent design and large-scale application and popularization of the high-performance heat-conducting electromagnetic shielding material.

The invention discloses a power grid intelligent power flow control method based on an improved cuckoo search algorithm. The method comprises the following steps: firstly, constructing a mathematical model of an AC-DC hybrid system, analyzing the characteristics and interaction of each device in the system, and determining a composite function taking the minimization of power generation cost, active loss and generator voltage deviation as targets according to an operation target; and then, decomposing the complex objective function into a plurality of sub-problems through hierarchical processing, and realizing respective solving. In hierarchical calculation, an improved cuckoo algorithm is adopted for optimization in a solution space, and an optimal power flow distribution and control strategy is obtained. And finally, intelligent layered power flow control of the alternating-current and direct-current hybrid system is realized according to an optimization result. According to the invention, the system response speed and the regulation efficiency are obviously improved, and the effect in the aspect of optimized operation is outstanding; compared with the traditional Cuckoo search CS algorithm and NSGA algorithm, the cost consumption is respectively saved by 2.21% and 4.49%; the experimental model is low in weekly average cost and high in power transmission efficiency, the power transmission efficiency is effectively improved, and the loss cost is reduced.

The invention relates to the technical field of dynamic response testing, in particular to a dynamic response testing method for an underwater model of a deep sea oil delivery hose. The method comprises the following steps: designing an experimental model of the deep sea oil delivery hose according to a similarity rate, calculating seabed boundary constraint conditions of the bottom end of the hose in an experiment, and equivalently representing the supporting effect of a vertical pipe on the hose in the horizontal direction and the vertical direction by adopting a spring stiffness model; based on technical parameters and sea conditions of the FPSO, hydrodynamic software is adopted to calculate and generate a six-degree-of-freedom movement time history of the robot; the six-degree-of-freedom movement process of the robot platform is input into a robot control system for loading testing; a camera is used for collecting images used for recording all the whole shooting and testing processes. According to the invention, the motion mode of the semi-submersible FPSO can be accurately loaded and reasonably simulated, and the high accuracy of the test result is ensured by optimizing the boundary condition of the hose model.

The invention discloses an underwater archaeological detection and identification simulation system based on an underwater cultural relic experiment model, and relates to the technical field of underwater archaeological detection.The underwater archaeological detection and identification simulation system comprises a detection management platform, and the detection management platform is in communication connection with the following modules: a cultural relic experiment model library building module, a cultural relic experiment model library building module and a cultural relic experiment model library building module, establishing a standardized and parameterized underwater cultural relic digital model library; and the multi-mode detection simulation module is used for simulating the response of acoustic, optical and electromagnetic detection technologies in the underwater environment. By constructing the standardized underwater cultural relic digital model library and combining geometry-material modeling and burying state configuration, accurate simulation of physical attributes and burying scenes of underwater cultural relics is realized, so that a simulation environment is closer to a real underwater condition, and the simulation efficiency is improved. The problems that traditional field detection depends on artificial diving, is low in efficiency and is limited by diving depth and time are effectively solved, and the trueness and reliability of detection simulation are remarkably improved.

The invention relates to the technical field of medical health, and provides an AD early warning model construction method and system based on a multi-dimensional test, and the method comprises the steps: collecting nerve image data, biomarker data and clinical feature data of a plurality of subjects, carrying out the abnormal data screening of a nerve image data set, carrying out the data cleaning of a biomarker data set, and carrying out the data cleaning of the biomarker data set; carrying out data coding on the clinical feature data set, respectively carrying out feature extraction on the preprocessed neural image data set and the preprocessed biomarker data set to obtain a neural image feature set and a biomarker feature set, and constructing a multi-dimensional feature matrix set; and obtaining a weighted multi-dimensional feature matrix set by using the training set matrix subset and the test set matrix subset, training the clinical prediction model by using the weighted multi-dimensional feature matrix set to obtain an early warning experiment model, and testing the early warning experiment model to obtain an early warning model. The problems of low diagnosis precision and abnormal data interference of a single data source can be solved.

The invention discloses an experiment design and supervision system and method, and belongs to the field of artificial intelligence. According to experiment requirements and sample characteristics, an experiment scheme is generated by adopting an experiment model and combining a local knowledge base and internet retrieval content; according to the experiment scheme, an experiment target is disassembled into specific equipment operation instructions, and instrument equipment and a robot are controlled to complete an action sequence; constructing a structured security rule base, processing unstructured information by using a large model, and checking whether the experimental scheme and the operation instruction violate security rules or not; the method comprises the steps of constructing a multi-field ethical rule base, extracting ethical related elements in an experiment scheme, and examining whether the ethical related elements conform to rules or not based on the ethical rule base, so that an experiment task is automatic and intelligent, the experiment efficiency is improved, and the experiment safety and compliance are guaranteed.

The method, device, equipment and storage medium for determining the surface wettability of a single mineral provided in the embodiments of the present application comprise: establishing a water molecule model, a crude oil component model and a mineral model with pores respectively; loading water molecules composed of the water molecule model and crude oil components composed of the crude oil component model into the pores of the mineral model respectively; optimizing a system model; performing molecular dynamics simulation on the optimized system model to obtain a mineral model with water molecules and crude oil components in the pores; calculating the interaction energy of the mineral and water and the interaction energy of the mineral and the crude oil components respectively; and determining the oil-water wettability of the mineral surface according to the interaction energy of the mineral and water and the interaction energy of the mineral and the crude oil components. In the embodiments of the present application, a measurement environment close to geological conditions can be created, and corresponding experimental models can be established for crude oil components and mineral components, which is conducive to characterizing the heterogeneity of reservoir wettability.

This invention relates to the field of surface subsidence and deformation technology, and particularly to a method for predicting abnormal surface deformation responses under the coupling of faults and loose layers. This method solves the technical problem of accurately assessing the surface subsidence and deformation patterns caused by coal mining under different geological conditions in existing technologies. The technical solution includes: establishing a three-dimensional numerical model using a Mohr-Coulomb constitutive model and setting the fault zone thickness; establishing eight sets of control experimental models and using the controlled variable method to study the effects of fault dip angle and loose layer thickness; performing simulation calculations by excavating the working face in 13 steps, with each step having a spacing of 30m; and effectively revealing the characteristics of how increasing fault dip angle promotes an increase in the maximum surface subsidence value, and how increasing loose layer thickness leads to an increase in the subsidence value followed by stabilization. This method can effectively assess the abnormal response patterns of surface subsidence and deformation under complex geological conditions.

The application discloses a high-speed wind tunnel applied to three-dimensional PIV, which comprises a nozzle section and an experimental section, wherein the experimental section is provided with an experimental area capable of placing an experimental model or a real aircraft; a concave cavity is arranged on the wall surface downstream of the experimental area of the experimental section, and an observation window is arranged on the wall surface of the concave cavity and faces the experimental area. The application is applied to the field of aerodynamics, and the angle of the observation window can be flexibly adjusted according to the characteristics of the supersonic flow field. The camera lens can be arranged vertically to the observation window during flow field observation, the imaging blur problem caused by the refraction of the observation window glass is avoided, clear flow field images are ensured to be shot, and the three-dimensional velocity field can be accurately calculated.

The application discloses a kind of simulation device and loading method for deep buried roadway slotting blasting in the technical field of experimental equipment, including double-shaft loading unit and the rock mass simulation test piece being arranged in the double-shaft loading equipment;First, make roadway experimental model, use electric blasting comprehensive experimental system to carry out blasting on the model, study the rock breaking mechanism of slotting hole blasting under different stress levels, the movement characteristics of post-blasting rock mass, dynamic strain evolution process and post-blasting effect analysis.This application mainly studies deep high-stress roadway, independently selects stress similarity constant and geometric similarity constant according to similarity experiment law, arranges blasting borehole and network according to on-site blasting live, which can realize comprehensive and real simulation of on-site slotting blasting process.Provides reliable experimental method and result for mine roadway slotting blasting mechanism research, which can provide guidance for on-site slotting blasting optimization.

The invention belongs to the technical field of heterogeneous traffic flow modeling and intelligent traffic management, and relates to a heterogeneous traffic flow macroscopic fundamental diagram experimental modeling method considering takeover behaviors, which comprises the following steps of: firstly, performing stage division on driving behaviors of L3-level man-machine co-driving vehicles, and selecting a car-following model; integrating TOV permeability and takeover rate, four typical car-following modes in heterogeneous traffic flow and a function degradation mechanism of cooperative adaptive cruise control to deduce theoretical scale models of vehicles of various models in the traffic flow. A joint simulation platform is built to restore a heterogeneous traffic flow scene, and a two-stage multi-cycle experiment process is designed to collect comprehensive traffic flow data under different density levels. Whether the experimental vehicle model proportion is matched with the theoretical vehicle model proportion or not is checked, and if yes, the improved Underwood model is used for fitting experimental data to obtain an experimental model of the macroscopic fundamental diagram. The method has the advantage that an accurate model basis is provided for heterogeneous traffic flow characteristic analysis and CAV cooperative control strategy formulation.

The application discloses a kind of overburden and surface fissure network reconstruction and inversion method for similar simulation experiment, it is related to network reconstruction and inversion technical field, including: carrying out simulation test, laying experimental model;Carrying out test, collecting three-dimensional point cloud data;Data is preprocessed, and feature vector is extracted;Based on feature vector reconstruction point cloud, get fissure feature, reconstruct fissure network;Optimize fissure network, carry out three-dimensional visualization and analysis;Model is imported into simulation software, iteratively optimized parameter, and the evolution process of fissure is inverted.The application uses three-dimensional point cloud imaging technology to realize high-precision data acquisition and processing, accurately extracts fissure feature;Based on fault image or point cloud data, three-dimensional reconstruction algorithm is used to reconstruct the three-dimensional model of fissure network, and multi-scale modeling is supported;Through numerical simulation, the evolution process of fissure is inverted, combined with machine learning algorithm to optimize inversion model, improve prediction accuracy.

This invention relates to an experimental data processing method, apparatus, AI analysis module, and computer equipment, belonging to the field of data processing. The method includes: segmenting multi-dimensional raw data by type and unifying the format; generating a similarity matrix based on temporal and spatial neighborhood information to mark abnormal fluctuation points; separating effective signals by matching a noise library with time-frequency features; extracting multi-layer features of basic statistics, time-series correlation, and trend changes; and dynamically selecting core features to update a tracking experimental model. The accompanying AI analysis module integrates hardware circuits for data segmentation, similarity calculation, anomaly marking, noise matching, feature extraction, and model updating, achieving full-process acceleration. This method significantly improves the efficiency of experimental data processing and the reliability of conclusions through multi-dimensional data compatible processing, accurate anomaly detection, multi-layer feature fusion, and adaptive model optimization, and is suitable for real-time analysis in various scenarios such as scientific research and industry.

The invention relates to a multi-phase flow closed-loop regulation and control experiment system for ventilation cavitation bubbles around a revolving body of a vertical water tunnel, and belongs to the field of cavitation water tunnel experiment technologies and equipment. Comprising a control system, a sensing system, a ventilation system, a rudder deflection system and an experimental model. The sensing system senses the state of the experimental model in real time; the control system is used for generating a rudder deflection control instruction and sending the rudder deflection control instruction to the rudder deflection system; a ventilation control instruction is generated and sent to a ventilation system; the rudder deflection system is used for adjusting rudder deflection parameters of the experimental model according to the rudder deflection control instruction, so that the attitude of the experimental model is adjusted, and the free pitching motion of the experimental model is realized; the ventilation system is used for adjusting ventilation parameters of the experimental model according to the ventilation control instruction so as to adjust the cavitation flow capacity of the experimental model; according to the invention, an experimental platform and a research basis are provided for research on flow phenomena such as cavitation falling and collapse and related large-scale and strongly-uncertain hydrodynamic characteristics under the flow direction gravity condition.

The utility model provides a kind of bridge substructure damage detection experimental model based on vehicle scanning method.The experimental model includes pier support system model, beam body model, vehicle model and self-sensing monitoring system.The beam body model is horizontally erected on two or three pier support system models, forming single-span simply supported beam or double-span continuous beam system.The vehicle model simulates driving on the bridge deck of beam body model.The pier support system model includes base model, pier model and support model from bottom to top.It can carry out multi-parameter control vehicle-bridge coupling vibration experimental model and its testing method.The model can monitor the stress state of bridge substructure in real time by introducing self-sensing monitoring system and intelligent damage simulation system, and automatically adjust damage parameters according to experimental requirements, improve the accuracy and repeatability of experiment.

The invention discloses an experiment table compiling system and method for a sanger sequencer, and the system comprises an experiment method: determining an experiment type, a dye set, a sequencing analysis model and experiment model information in an established experiment method, binding and summarizing the experiment model, the dye set and the sequencing analysis model, and forming an experiment method; sample file names and sample file storage position information are determined in the established sample files, and the sample file names and the sample file storage positions are bound and summarized to form one sample file; according to the kit, an analysis method is defined in the established kit, Panel and interior label information is analyzed, and the analysis method, the Panel and the interior label are bound and summarized to form the kit; and a runtime library: binding and summarizing the established experimental methods, sample files and kits to form the runtime library. According to the invention, the user only needs to select the runtime library for one time to complete the tabulation processing of the experiment, so that errors caused by selection are reduced.

The invention belongs to the technical field of wind load experiment testing, and particularly discloses a modularized equivalent wind load experiment testing system which is characterized by comprising a wind load simulation module, a data acquisition and processing module, a model fixing and supporting module, a control system module and a function extension module. The wind load simulation module is used for adjusting wind load strength and wind speed and realizing 0-360-degree wind direction adjustment; the data acquisition and processing module is used for monitoring physical quantities such as displacement and strain of the experimental model and wind speed, wind pressure and the like of a wind field, and processing to generate an experimental result and an analysis report; the model fixing and supporting module is used for fixing the experimental model and adjusting the levelness and perpendicularity of the experimental model; the control system module is used for setting experiment parameters, controlling experiment start and stop and process monitoring and realizing data transmission between modules; the function extension module is used for measuring wind pressure distribution on the surface of the model, displaying the flow state of a wind field and monitoring vibration response of the model; all the modules are connected and transmit data through standard interfaces.

The invention discloses an experimental method and system for verifying multi-scene validity of a large power language model. The method comprises the following steps: constructing a layered experiment model for simulating a full-link business process according to a power business scene requirement; performing a contrast test on the experimental group and at least one control group, collecting data, calculating experimental data to obtain multiple indexes, calculating subjective and objective weights based on expert experience and data volatility, and fusing the subjective and objective weights based on an AHP-entropy weight method to obtain a plurality of indexes; therefore, a dynamic evaluation mechanism adaptive to different power scene priority changes is realized. And iterating the model through a dynamic evaluation result so as to carry out term recognition error analysis and knowledge graph incremental training, thereby realizing data-model closed-loop optimization. According to the method, verification scientificity and conclusion credibility are remarkably improved, subjective experience of experts and objective volatility of data are effectively fused, the stiffness defect of traditional static indexes is overcome, and direct driving of evaluation results on model iteration is achieved.

The utility model provides a test model device for testing a load transfer effect of a fill close to a bridge foundation, which comprises a first support plate, a first through groove is arranged on the first support plate, the bridge foundation is arranged in the first through groove, and a bridge is arranged at the top end of the bridge foundation; a third supporting plate is arranged on one side of the first through groove and located at one end of the first supporting plate, a plurality of rib supporting plates are arranged on the third supporting plate, a pressure bearing plate is arranged above the rib supporting plates, a pressurizing assembly is arranged above the pressure bearing plate, slope ratio sliding plates are arranged on the two sides of the pressurizing assembly, and a slope ratio sliding plate is arranged on the slope ratio sliding plates. A connecting plate is arranged between the slope ratio sliding plates, and second through grooves are formed in the slope ratio sliding plates. By adopting the test model device for testing the load transfer effect of the fill close to the bridge foundation, the function of controlling the influence coefficient of the fill pressure behind the support structure under different factors on the close bridge foundation is realized, and the original slope ratio and the distance from the bridge foundation to the support structure can be adjusted.

The invention relates to the technical field of ocean engineering tests, in particular to an ultra-large floating platform laboratory model attitude control device. The device comprises a rotary telescopic mechanism, a fixing device and a base, the rotary telescopic mechanism comprises a turbine worm mechanism, a servo electric cylinder and a servo motor; the worm and gear mechanism and the servo motor are both installed on the base, and the output end of the servo motor is connected with a worm of the worm and gear mechanism. The servo electric cylinder is installed on the wheel face of a worm wheel of the worm and gear mechanism, and the output end of the servo electric cylinder downwards penetrates through the worm wheel of the worm and gear mechanism and the base in sequence and is connected with the fixing device. Accurate adjustment of the vertical position of the experimental model can be realized through the servo electric cylinder, flexible adjustment of the horizontal angle is realized through cooperation of the worm and gear mechanism and the servo motor, the working conditions of the floating structure for coping with waves coming in different directions and different vertical postures can be accurately reproduced, and the multi-scene experimental requirements are met.

The invention belongs to the technical field of animal model construction, and particularly relates to a spinal cord hypertension syndrome disease model modeling method and application. A guinea pig or a macaque is used as an experimental animal, and a composite biostimulation material and an accurate pressure regulation and control modeling system are combined to realize modeling. The composite biostimulation material is composed of a core pressure slow-release layer, a middle biocompatible coating and a surface targeting binding layer, has excellent biocompatibility, degradability and targeting property, and can reduce non-specific inflammatory response; the precise pressure regulation and control modeling system comprises a pressure sensing module, a micro-injection module, a real-time monitoring module and a main control module, and can dynamically regulate and control the pressure in the spinal cord in a closed-loop mode. The modeling method is easy and convenient to operate and good in repeatability, the model building success rate exceeds 95%, the pathological features are highly similar to clinical features and can last for 4 weeks, a reliable experimental model is provided for mechanism research, drug screening and therapeutic schedule evaluation of the spinal cord high pressure syndrome, and the modeling method is suitable for large-scale popularization.

This invention discloses a modeling method for predicting cucumber photosynthetic rate using a BP (Backpropagation) model that integrates transpiration rate, relating to the fields of smart agriculture and machine learning. The method includes: continuously monitoring the photosynthetic physiology of cucumber plants in the fruiting stage under greenhouse cultivation conditions to obtain photosynthetic monitoring characteristic data; normalizing the photosynthetic monitoring data and dividing it into training and testing sets; constructing an initial BP neural network as an experimental model, using the coefficient of determination, adjusted coefficient of determination, and root mean square error as evaluation indicators to determine the optimal transfer function combination of the hidden and output layers of the experimental model; optimizing the experimental model using Bayesian optimization and five-fold cross-validation with the root mean square error as the optimization objective to determine the optimal hyperparameter combination of the experimental model; and remodeling the experimental model based on the optimal transfer function combination and the optimal hyperparameter combination to obtain a cucumber BP photosynthetic rate prediction model integrating transpiration rate.

A storable mutual work theorem verification experiment device belongs to the technical field of material mechanics practical teaching and comprises an assembled storage assembly, an assembled experiment model assembly and a measuring and loading assembly. The assembled storage assembly serves as a part storage box of the assembled experimental model assembly in a storage state, and serves as a supporting foundation of the assembled experimental model assembly in a non-storage state; the measuring and loading assembly is a common experimental apparatus in a laboratory, is not stored by the assembled storage assembly, and is only used in cooperation with the assembled experimental model assembly in an assembled state during an experiment. According to the invention, complete verification of the interoperability theorem of work can be realized, and the blank of verification of the generalized load by the existing related experimental device is filled; the experimental device has the structural characteristic of being convenient to store, after the experimental device is used, the experimental device can be disassembled into parts to be integrally stored, the occupied space of the experimental device in an idle state is greatly compressed, and the space utilization rate of a laboratory is effectively increased.

The invention relates to a product function reliability domain experimental design method, belongs to the technical field of reliability engineering, and solves the problem that an existing reliability test technology cannot comprehensively recognize the function reliability domain of a product and the reliability level of the product in the function reliability domain, and further embodies in engineering practice. And when the test result shows that the reliability of the current product cannot meet the requirement, the information of how to carry out design improvement on the product is difficult to provide based on the result obtained by the test. The method comprises the following steps: constructing a product function reliability domain experimental model; determining a control variable, an intervention variable and an observation variable of a function reliability domain experiment based on the function reliability domain experiment model; determining an intervention variable constraint, a sample size constraint and an experiment time constraint of a product performance degradation law experiment; and designing a product performance degradation law experiment scheme based on the determined constraint.

The application relates to the technical field of medical health, and provides an AD early warning model construction method and system based on multi-dimensional testing, which comprises the following steps: collecting neural image data, biomarker data and clinical feature data of a plurality of subjects, screening abnormal data of the neural image data set, performing data cleaning on the biomarker data set, performing data coding on the clinical feature data set, respectively extracting features from the pretreated neural image data set and the pretreated biomarker data set to obtain a neural image feature set and a biomarker feature set, constructing a multi-dimensional feature matrix set, obtaining a weighted multi-dimensional feature matrix set by using a training set matrix subset and a test set matrix subset, training a clinical prediction model by using the weighted multi-dimensional feature matrix set, obtaining an early warning experimental model, and testing the early warning experimental model to obtain an early warning model. The application can solve the problems of low diagnosis precision of a single data source and abnormal data interference.

Embodiments of the present application disclose a multi-layer fracturing simulation experiment method, device, electronic equipment and medium. The method comprises: constructing a multi-layer fracturing simulation experiment model and at least three fracturing layers; sequentially installing the at least three fracturing layers into the constructed multi-layer fracturing simulation experiment model for fracturing, and determining the crack state of the at least three fracturing layers in the fracturing process under different fracturing conditions; wherein the fracturing conditions include injection pressure and injection flow rate; corresponding the crack state with the fracturing conditions, determining the target fracturing condition and crack extension law of the at least three fracturing layers when the crack initiates. By using the technical scheme of the embodiments of the present application, the crack extension of each fracturing layer under different fracturing conditions is simulated and analyzed, the target fracturing condition and crack extension law of the crack extending through the layer are obtained, and guidance is provided for well and layer selection and segmented fracturing construction optimization design.

The invention discloses a die cutting process analysis method based on LS-DYNA software, and relates to the technical field of analogue simulation methods, and the die cutting process analysis method comprises the following steps: step 1, establishing a two-dimensional simplified model of a cutter and a cut material by using HyperMesh software; step 2, dividing grids, encrypting a processing area, and prefabricating a node strain failure criterion for a material area which may fail; 3, defining a material; 4, setting the unit type as a plane stress shell unit; 5, setting boundary conditions, and constraining the gasket; step 6, defining a contact relationship, and adopting automatic surface-to-surface contact between the material layers; and 7, setting control parameters, and outputting k file analysis. According to the invention, the test cost is reduced, the test model and the test environment are more convenient to build, digital simulation and analysis of die cutting processing are realized, and the analysis result is more effective.

The invention relates to a centrifugal model experiment device for simulating multiple dynamic water effects. The centrifugal model experiment device comprises an experiment model box, an experiment water tank, a folding fluid regulator, an internal circulation water supplementing pipeline, a water vapor adjusting pipeline and a reservoir water pipeline. The experimental water tank is arranged in the experimental model box; a side slope generalization model is arranged outside the top surface of the experimental water tank, and the folding fluid regulators are arranged above and on the side surface of the side slope generalization model; one end of the internal circulation water replenishing pipeline is connected with the experimental water tank, and the other end is connected with a water source and the folding fluid regulator; one end of the water-gas adjusting pipeline is connected with the experimental water tank, and the other end extends out of the experimental water tank; one end of the reservoir water pipeline extends to the bottom surface of the experimental water tank, and the other end extends out of the top surface of the experimental water tank. Compared with the prior art, the device has the advantages that the foldable fluid regulator can simulate a rainfall working condition and an underground water seepage working condition; water in the experiment water tank is controlled to flow into the experiment model box through the reservoir water pipeline, and the reservoir water level lifting working condition can be simulated.

This invention discloses a high-temperature and high-pressure micro-displacement method based on an isoporosity dual-pore microfluidic model. Addressing the issues of irreversible diagenesis and the difficulty in obtaining paired samples before and after evolution, a baseline model A and an experimental model B are constructed. Isoporosity constraints are used to eliminate interference from reservoir property differences, enabling a quantitative evaluation of the influence of dissolution pore structure on seepage patterns. The method includes: S1, extracting the primary intergranular pore skeleton and secondary dissolution pore characteristics from typical core thin sections of the same stratum; S2, constructing the baseline model A; S3, embedding dissolution pore units within its topological space and using an equivalent volume compensation algorithm to adaptively scale the intergranular channels to form model B; S4, verifying the porosity accuracy of the completed baseline model A and experimental model to ensure that the pore volume deviation between the two models does not exceed 5%; S5, conducting comparative displacement experiments under real high-temperature and high-pressure conditions.