47results about How to "Reliable basis for decision-making" patented technology

The invention discloses a hydroelectric generating set upper guide shoe temperature intelligent prediction method based on multi-source data fusion, and relates to the technical field of hydroelectric generating set state monitoring and intelligent early warning, and the method comprises the steps: collecting and preprocessing multi-source operation data of a hydroelectric generating set, and constructing a unified time sequence data set; according to a preset working condition zoning rule, the operation working condition type of the current hydroelectric generating set is judged, and a working condition label corresponding to the thermal state of the upper guide shoe is obtained; performing feature importance sorting on the multi-source fusion feature set to obtain an optimal feature subset; on the basis of the optimal feature subset, an upper guide shoe temperature integrated prediction model is established and trained; and predicting an upper guide shoe temperature prediction value, and judging whether a temperature abnormal rise risk exists or not. According to the method, high-precision prediction and real-time abnormal trend identification of the temperature of the guide shoe on the hydroelectric generating set are realized through working condition classification, feature optimization and integrated prediction strategies, and meanwhile, reliable data support and decision basis are provided for intelligent operation and maintenance.

The invention discloses a multi-modal disambiguation and knowledge graph embedding-based power grid Internet of Things equipment plug and play method and system, and the method comprises the steps: analyzing an ambiguous user instruction through a multi-modal module, and fusing context and sensor data to generate a clear configuration intention; a knowledge graph containing equipment configuration and topological resources is constructed, after the knowledge graph is converted into a vector through embedding, an optimal access scheme and a physical model are matched through machine learning, meanwhile, power grid operation rule verification is conducted on a matching result, and it is ensured that the scheme can be directly executed. According to the method, the problems of inconsistent object model matching, manual configuration migration, weak ambiguous instruction analysis and the like are solved, automatic access of novel switches, photovoltaic equipment and other equipment is realized, and the automation level of the power grid internet of things is improved.

The application provides a posture control method and system of a folding wing unmanned aerial vehicle, and relates to the technical field of aircraft control. Firstly, the application collects pressure information and strain information of the folding wing unmanned aerial vehicle, and determines whether the wing has aerodynamic-elastic coupling during the action according to the two kinds of information. Then, when it is determined that the aerodynamic-elastic coupling occurs, the airflow control instruction corresponding to the current coupling state is determined. Then, according to the airflow control instruction, the asymmetric additional airflow on the wing surface is generated by controlling the pulse jet execution mechanism. Finally, the compensation moment is generated by the additional airflow to offset the body posture disturbance caused by the target action of the folding wing unmanned aerial vehicle. The technical scheme provided by the application not only realizes the real-time perception, accurate judgment and active inhibition of the nonlinear aerodynamic-elastic coupling disturbance of the folding wing unmanned aerial vehicle during the shape conversion process, but also improves the stability of the flight posture of the folding wing unmanned aerial vehicle.

The invention provides a method and system for accurately controlling the temperature of an adhesive production reaction kettle, and relates to the technical field of adhesive production, and the method comprises the steps: applying a transient heating pulse to the reaction kettle in adhesive production, collecting the surface temperature data of a container wall, processing the data, calculating a temperature decline curve, and extracting key information of heat transfer characteristics. A heat abnormal area caused by uneven mixing is identified based on spatial distribution of key information, and the thermal state change of the adhesive reaction degree is estimated in combination with a time change mode. And adjusting a heating pulse mode and a working state of the stirring device according to the abnormal area and the change mode, and introducing a pause detection stage to collect surface temperature feedback under weak heating pulses so as to confirm an adjustment effect. And finally, associating the key information with the reaction degree based on feedback to generate a temperature adjusting command, and adjusting the heating or cooling intensity, so that accurate control on the temperature of the reaction kettle can be realized, and the production quality and efficiency are improved.

The application provides a sea area efficiency evaluation method and system based on geospatial analysis technology, which comprises the following steps: acquiring multi-source evaluation indexes of each unit, calculating first-order statistical moments, calculating spatial Gini coefficients and local information entropy of each index domain, correcting a preset resolution coefficient by using the local information entropy when the Gini coefficient exceeds a threshold, calculating deviation sequences of each unit index value and neighborhood first-order statistical moments, setting an asymmetric processing function according to the local information entropy, amplifying the absolute value of a negative deviation by using an exponential function, and compressing the value of a positive deviation by using a logarithmic function, wherein the local information entropy is positively correlated with the base number of the function, calculating an initial efficiency value based on the processed deviation sequences and the corrected resolution coefficient by using a comprehensive grey correlation degree, calculating a local Moran index based on the initial efficiency value to determine a spatial clustering type, correcting the resolution coefficient twice according to the spatial clustering type, recalculating the comprehensive grey correlation degree, and obtaining a sea area efficiency evaluation result.

The invention discloses a power distribution network line environment meteorological data prediction method and system, and belongs to the technical field of meteorological prediction, and the method comprises the steps: S1, carrying out the space-time alignment and quality control of collected historical environment meteorological data of a power distribution network line; s2, taking the lightweight long-short-term memory network as a learning framework of a time sequence evolution rule in the historical environmental meteorological data, and taking the physical constraint as a learning constraint of the time sequence evolution rule in the historical environmental meteorological data to construct a hybrid prediction model; and S3, inputting the historical environmental meteorological data into the hybrid prediction model for multi-scale space-time prediction to obtain an environmental meteorological field of each interval of the power distribution network line, obtaining a confidence interval of each interval according to the environmental meteorological field, and obtaining predicted environmental meteorological data of each interval according to the confidence interval. The technical problem that the environmental weather prediction accuracy is difficult to improve in the prior art is solved.

The present application relates to geological exploration and three-dimensional modeling technical field, disclose a kind of fluorite ore three-dimensional exploration modeling and analysis system, the system includes: data preprocessing and metallogenic period definition module, space-time geological structure tensor field construction and dynamic updating module, dynamic geodesic distance field calculation module, and multi-stage feedback driven level set modeling module.The present application is evolved by executing multi-stage time sequence, the model boundary of previous geological period evolution convergence is used as the starting boundary of current period, and the inheritance type mineralization is simulated.In evolution, the system also performs dynamic feedback: narrow band area near the current evolution surface is identified, the tensor field module is instructed to update locally in the narrow band, and the geodesic distance field is recalculated to guide the next evolution.The present application can also generate a three-dimensional boundary probability volume by applying a disturbance to the input parameters and repeating the modeling to quantify the uncertainty of the model.

The application discloses a kind of whole life cycle stability evaluation methods of cutting embankment slope, it is related to geotechnical engineering and slope stability evaluation technical field.The method includes: step one, using survey multi-source geological information to establish parameter probability distribution model;Step two, establish construction process numerical model, carry out pre-reliability evaluation;Step three, based on actual disclosure information, condition probability correction is carried out to parameter probability distribution;Step four, based on posterior distribution, recalculate reliability, dynamically optimize construction scheme;Step five, after construction, monitoring system is arranged;Step six, using monitoring data sequential update model state and parameter;Step seven, based on calibration model, predict time-varying reliability during operation under extreme working condition;Step eight, set graded early warning threshold, realize reliability and monitoring joint early warning;Step nine, according to early warning decision maintenance scheme, and feedback model after maintenance, cycle executes step six to step nine.

The application discloses a kind of sepsis multiple complications prediction method, system, storage medium and equipment, belong to intelligent medical technical field, method includes the following steps: S1. acquisition sepsis patient multidimensional time series physiological data;S2. medical perception feature vector is constructed to the time series physiological data;S3. sepsis multiple complications prediction model is constructed and introduced physiological logic space-time composite mask matrix;S4. the output multiple complications prediction probability using hierarchical expert hybrid network H-MoE;S5. the model is trained using dynamic weighted loss function based on task uncertainty;S6. risk output and feature contribution degree explanation;The model of the application is introduced by physiological logic mask and H-MoE architecture, realizes the accurate, synchronous prediction of sepsis complex complications, can effectively filter redundant features, and reveal the complex correlation therebetween, provide quantitative decision support for clinician, to realize early intervention and individualized treatment.

The invention discloses a bearing residual service life prediction method based on a relevance vector machine and an NNHTS model, and the method comprises the steps: firstly extracting the time domain features of a bearing vibration signal, and constructing a comprehensive health index through filtering noise reduction and principal component analysis; secondly, optimizing variational mode decomposition parameters by using a genetic algorithm, and decomposing the health index sequence into high-frequency and low-frequency components; then, training and predicting the high-frequency component and the low-frequency component by adopting an NNHTS model and a correlation vector machine respectively; and finally, superposing the predicted values of the components, and reconstructing a health index decline curve to predict the residual life. According to the method, through decomposition, hierarchical modeling and integration strategies, the fitting ability of deep learning to complex fluctuation and the sparse probability modeling advantage of a relevance vector machine to trend are combined, and the prediction precision and working condition adaptability are remarkably improved.

A method for determining the association and continuity combination of top-side-rock structural surfaces based on unsupervised learning is proposed. This method involves investigating the structural surfaces of the roof and side-rock masses in underground engineering projects to obtain geometric and statistical information. A unified encoding is performed to construct a joint feature set of roof-side-rock structural surfaces. An unsupervised density-based clustering algorithm is used for initial cluster analysis of the joint feature set. A swarm intelligence-based Lüperfox optimization algorithm is introduced to optimize key parameters of the DBSCAN clustering algorithm. A Latin hypercube sampling method is used to improve the population initialization process of the Lüperfox optimization algorithm, resulting in better DBSCAN hyperparameters and optimized structural surface clustering results. The method determines the structural surface system affiliation of structural surfaces in different locations, quantitatively calculates the spatial continuity and combination relationship of structural surfaces, and outputs the spatial association determination results and continuity levels of the roof-side-rock structural surfaces. This method enables intelligent identification and quantitative analysis of the spatial relationship between roof and side-rock rock structural surfaces.

This invention belongs to the field of mine safety monitoring technology, specifically involving an intelligent management system based on miners' positioning data processing. The system includes a wearable precision inertial measurement device, a multi-source data fusion gateway, an underground wireless transmission network, a biomechanical feature extraction server, a behavioral quality evaluation management platform, and a multi-dimensional early warning terminal. The wearable device collects high-frequency motion vector data of miners in real time, performs preliminary noise reduction using the gateway, and the server establishes a gait cycle segmentation model to accurately extract gait feature parameters such as stride length, ground contact time interval, and body center of gravity oscillation frequency. This invention realizes a shift from simple spatial location monitoring to in-depth monitoring of biological signals and behavioral quality, filling the monitoring blind spots caused by passive displacement and improving the proactiveness of mine risk early warning and the accuracy of safety supervision.

The application discloses a kind of cable health state evaluation method and system based on single-end open circuit impedance spectrum, belong to nondestructive testing and impedance spectrum analysis technical field.This method only needs to carry out frequency sweep test at one end of cable and keep open circuit at the end, after obtaining wideband input impedance spectrum, extract real part data and identify resonance extreme point;The upper and lower envelope lines are constructed by extreme point interpolation, and the high-frequency characteristic impedance and attenuation constant of the full frequency band are derived;Further, the distributed parameters of the cable are calculated by combining the phase constant.By comparing the characteristics of the distributed parameters, the health status of the cable can be accurately evaluated, and the physical mechanism can effectively distinguish between thermal aging and moisture damage.The application eliminates the cumbersome of double-end cooperation test, does not need to intercept samples, greatly saves the test time, has wide applicability and strong anti-interference ability.

The invention discloses a corn pathogen early warning method and system based on big data, and relates to the technical field of agricultural early warning, the method comprises the following steps: comparing sequences in historical small RNA deep sequencing data and verified viral pathogen monitoring data, and establishing a mapping relationship between the two to obtain a pathogen-small RNA reference corresponding relationship library; matching the small RNA deep sequencing data of the current monitoring period with the pathogen-small RNA reference corresponding relationship library, and extracting sequence fragments consistent with the mapping relationship in the library to obtain a target viral pathogen feature sequence; coupling the target viral pathogen feature sequence with the current environmental factor data; according to the method, the core value in the multi-source data can be effectively mined, the early infection signal of the pathogen can be accurately captured, the early accurate prediction of the transmission risk and potential outbreak of the corn viral pathogen is realized, a timely and reliable decision basis is provided for field prevention and control, and the corn production safety is guaranteed.

This invention discloses a method and system for steer-by-wire road feel tracking control of commercial vehicles. The method includes: executing multiple control cycles; within each control cycle: acquiring steering state parameters and vehicle motion state parameters during vehicle operation, and reading the estimated road friction capability and feedback torque from the previous control cycle; generating the current road feel target torque; solving the nonlinear state-space model to obtain the estimated road friction capability for the current control cycle; setting the length of the prediction time domain and the control law parameters of the error feedback controller; adaptively allocating the output weights of the two-layer controller, correcting the initial control command, and generating the final control command; and outputting the control command of the current control cycle to the road feel actuator to acquire the feedback torque. This invention can improve the road feel tracking accuracy and operating condition adaptability of commercial vehicles.

This invention relates to the field of power distribution protection technology, specifically disclosing a gas leakage monitoring device for power distribution equipment suitable for extremely cold environments. The device includes: a data acquisition unit comprising multiple gas sensor probes deployed at vertical heights within the monitored area; a data processing unit that performs temperature compensation correction on the gas concentration signal to obtain corrected discrete-point concentration data; a fusion calculation unit that reconstructs the spatial concentration field of the monitored area based on the corrected discrete-point concentration data, performs dynamic data assimilation on the spatial concentration field to obtain an optimal estimated concentration field, calculates at least two characteristic parameters characterizing the gas state based on the optimal estimated concentration field, and outputs a decision output result; and a result application unit that identifies and warns of leakage status based on the decision output result. This invention can avoid false alarms and missed alarms caused by gas liquefaction or stratification, making it suitable for gas leakage monitoring of power distribution equipment in extremely cold environments.

The invention relates to a crane health state real-time monitoring method and system, and the method comprises the steps: obtaining original operation data of a crane, and carrying out the calculation to obtain a time domain feature and a frequency domain feature; a structure including a physical topological graph is constructed, node weights are obtained through calculation according to the physical topological graph structure, feature fusion is carried out on the time domain features, the frequency domain features and the node weights, and a fault diagnosis result is generated in combination with the spectral analysis graph; a life prediction model is constructed, and pre-training and fine tuning are carried out; and inputting the real-time time domain feature and the frequency domain feature into the fine-tuned service life prediction model, and outputting the remaining service life. According to the method, the fault diagnosis precision is improved, the actual crane structure can be better adapted, and the generalization ability of the model is improved.

The application discloses a kind of based on knowledge graph and big model fusion's fishery science and technology project innovativeness evaluation method, the method is as follows: S1: from English literature database and patent database automatically collects fishery science and technology related paper and patent data, constructs multi-source heterogeneous dataset;S2: to multi-source data is de-duplication, cleaning and standardization processing, through word segmentation, word class restoration and feature extraction form unified structured data representation;S3: based on structured data carries out entity, relationship and attribute extraction, combines cross-language alignment and expert knowledge verification constructs fishery science and technology field knowledge graph;S4: to project research content and existing technology carries out semantic correlation matching and difference analysis;S5: constructs and includes novelty and frontiers innovativeness evaluation model, carries out quantitative evaluation to project and exports evaluation result.The method not only improves the accuracy and systematization level of innovation point identification, but also guarantees the traceability and comparability of evaluation result.

The application relates to the cross field of artificial intelligence and environmental health, and discloses a membrane nephropathy prognosis risk stratification method based on exposure spectrum and clinical indicators. First, demographic information, clinical detection indicators and various environmental pollutant exposure data of patients are collected to construct a multi-level exposure spectrum feature, including single pollutant concentration, same type comprehensive exposure factor and pollutant synergistic interaction term. Then, the exposure spectrum and the clinical indicators are fused, various machine learning algorithms are used to construct candidate prediction models, the optimal prognosis risk stratification model is screened through cross validation and multiple performance indicators, and the SHAP value is used to quantify the contribution degree of each feature. Further, a double machine learning method is used to control confounding variables and estimate the causal effect of key pollutants on prognosis. On this basis, individual risk stratification and main risk driving factor identification of membrane nephropathy patients are realized, and scientific basis and guidance are provided for clinical decision making.

The application relates to a satellite energy prediction method and device, computer equipment and a storage medium. State data and task plan data of a target satellite are acquired, a pre-trained hybrid prediction model is called based on the state data, a prediction SOC sequence of the target satellite in a prediction period is obtained, the remaining battery capacity in the prediction period is determined according to the prediction SOC sequence, and an energy budget value for load task planning is calculated according to the remaining battery capacity and the task plan data. Since the hybrid prediction model integrates ARIMA, CNN and LSTM, the advantages of different algorithms are fully utilized through structured division of labor and cooperation, so that the energy state and budget of the satellite in a complex and dynamic environment can be accurately predicted, reliable and forward-looking decision basis is provided for intelligent management of satellite energy, and long-term reliable operation and maximum service efficiency of the satellite can be ensured.

The application discloses a straw distributed energy supply regulation and control method and system for supply and demand balance and a storage medium, and belongs to the technical field of biomass distributed energy supply regulation and control. Multidimensional data of the energy supply side and the energy use side are collected, a double prediction model fusing time sequence characteristics and correlation characteristics is constructed, and accurate prediction of energy use load and energy supply capacity is realized. A straw energy supply and use balance matrix is established, the predicted curve is compared in real time to identify the supply and demand deviation type, and a dynamic regulation and control strategy of load increment, load peak clipping and energy storage scheduling is generated based on the deviation. Actual operation data after regulation and control is collected, load matching degree deviation is calculated, and abnormal samples are marked, and the double prediction model parameters are iteratively optimized through a gradient descent algorithm. The application improves the supply and demand matching accuracy and operation stability of the straw energy supply system, adapts to the dynamic regulation and control demand of complex scenes, and solves the problems of inaccurate regulation and control mode prediction and low matching degree.

The invention discloses a high temperature-low temperature prediction method and device based on numerical weather forecast, and the method comprises the steps: obtaining the original data of a numerical weather forecast mode in a prediction region, collecting the historical air temperature observation data of a plurality of meteorological stations in the prediction region, and enabling the historical air temperature observation data to comprise the daily maximum air temperature and the daily minimum air temperature; performing quality monitoring on the collected numerical weather forecast data, and sorting and cleaning the original data and the historical air temperature observation data; constructing sample features based on the preprocessed data, marking a true value, taking the sample features as input features, and training an established high-temperature-low-temperature prediction model; and carrying out the same preprocessing step as the training data on numerical weather forecast data in a prediction area obtained in real time, and inputting the data into the trained prediction model to obtain an air temperature prediction value of the daily maximum air temperature or the daily minimum air temperature. The prediction result is closer to the actual temperature change trend, and a more reliable decision basis is provided for each industry.

This application discloses a digital management method and system for hospital nursing quality indicators, relating to the field of medical and nursing quality information technology. The method includes the following steps: setting assessment rules for evaluating nursing process quality indicators; acquiring patient nursing records and constructing a patient-centered event time series; based on the assessment rules, filtering nursing events related to the corresponding nursing process from the event time series; dynamically adjusting the strictness of the assessment rules based on the identity information of nursing staff and the clinical status information of patients; verifying the event time series according to the adjusted assessment rules and obtaining verification results; determining the compliance of the nursing process based on the verification results, and generating a compliance conclusion and audit report for the nursing process. This application can intelligently distinguish between genuine nursing quality defects and discrepancies in record appearances caused by differences in personnel and context, solving the problems of fragmented processes, isolated data, and delayed assessment in traditional nursing quality management.

The present application relates to the field of welding robots, and aims to solve the problem that conventional welding control systems lack real-time perception and adaptive ability to dynamic changes in the welding process, and cannot realize real-time evaluation, early warning and collaborative supervision, and specifically provides an automatic welding robot intelligent control system, which realizes precise progress management and work hour prediction of the whole welding process by collecting welding feedback information in real time, comparing with standard welding conditions, dynamically estimating the remaining welding time through a self-learning speed prediction model, combining a multi-dimensional, threshold-based dynamic sampling comparison mechanism to compare the welding path, thickness and heat output with high precision, and finally obtaining comprehensive error analysis, which can not only reflect the welding quality in real time, but also evaluate the severity of quality problems, provide accurate and reliable decision basis for whether to interrupt the work or adjust the process, effectively avoid the misjudgment of traditional single parameter alarm, and realize online quality control.

The invention belongs to the technical field of valve sealing performance testing, aims to solve the problems that in the prior art, microscopic symptoms of flow state transition cannot be captured in the extremely early stage of leakage development, and the essential type of a fault cannot be further deconstructed from the coupling relation of multi-source data, and particularly relates to a valve sealing performance testing system based on multi-source data fusion. Comprising an operation parameter acquisition module, a dynamic feature extraction module, a mapping analysis module, a distortion analysis module and a state evaluation module. According to the method, a reference coordinate system with a pressure zero crossing point as a time anchor point is constructed, a flow zero crossing point is mapped into the coordinate system, and traditional macroscopic magnitude comparison is converted into microstructure analysis of a pressure-flow coupling relation; the fundamental transformation of leakage early warning from'macroscopic threshold triggering 'to'microstructure distortion capturing' is realized, and the timeliness of early detection of leakage is remarkably improved.

The application provides a contraband detection equipment risk early warning method and system using a cloud network end, wherein Raman spectrum data of a sample to be detected is obtained and wavelet transform is performed to generate a feature spectrum to be matched; a cloud server gathers non-alarm spectrum data uploaded by multiple terminals, and generates an environment background feature spectrum set by clustering; the feature spectrum to be matched is matched with a standard spectrum library and a background spectrum set, and based on spectrum cross-correlation, peak shift and peak shape correlation weighted operation, a comprehensive matching score is obtained; when the comprehensive matching score of the feature spectrum to be matched and any background spectrum is higher than a background suppression threshold, an adjusted contraband comprehensive matching score is generated; based on the adjusted contraband comprehensive matching score, a ratio of the highest score and the second highest score is calculated to determine a matching ambiguity coefficient, and in combination with the highest score, a preset threat level of the corresponding contraband and the matching ambiguity coefficient, a risk early warning level is output.

This invention provides a method and apparatus for evaluating the credibility of generated content, relating to the fields of computer science and artificial intelligence. The method includes: collecting information sources from multiple platforms; extracting features from the information sources to obtain multi-dimensional features; defining graph nodes and edges, assigning weights, and learning graph neural networks from the multi-dimensional features to obtain a directed weighted information source authority graph; performing end-to-end self-supervised learning on the information source authority graph using a graph convolutional network or graph attention network to obtain the graph neural network authority score for each node; performing conflict graph modeling and confidence propagation iterative algorithm processing on the information source authority graph to obtain a confidence propagation score; setting a timeliness factor, and weightedly fusing the graph neural network authority score, confidence propagation score, and timeliness factor to obtain a credibility score. This method achieves dynamic, accurate, and multi-platform collaborative evaluation of information source credibility.

The invention provides a microwave reaction cavity heating uniformity analysis method and system, and relates to the technical field of microwave heating. Real-time microwave excitation and cavity wall temperature array signals are synchronously collected, and excitation weights are analyzed by using a pre-built model and a regularization inversion algorithm; on-line and dynamic analysis of the heating state of the microwave reaction cavity in operation is achieved, the hysteresis problem of traditional detection depending on shutdown is solved, random noise can be effectively stripped from an actually measured temperature field through screening of dominant modes, construction of energy space distribution and weighted fusion, and the detection accuracy is improved. The method comprises the following steps: firstly, extracting a characteristic temperature field dominated by the inherent electromagnetic characteristics of a cavity, enhancing the physical mechanism property and anti-interference capability of analysis, then introducing compensation calculation based on the dielectric temperature change characteristics of the material, correcting heat distribution distortion caused by the change of the attributes of the material, and calculating the heat distribution distortion; and the obtained reference temperature field can more truly reflect the heating uniformity capability of the cavity.

The invention provides a complex mountainous area landslide rapid identification method and system based on multi-source data, and relates to the technical field of complex mountainous area landslide rapid identification, through a deep mining and attention weight fusion algorithm, the influence weight of each reference key influence attribute is calculated, the weight directly represents the confidence coefficient of the attribute, and the confidence coefficient of each reference key influence attribute is calculated. And screening out core attributes which have obvious influence on the landslide. And on the basis of the effective reference key influence attributes and the corresponding influence weights, target key influence attribute description content is generated through multi-analysis-unit cooperative processing and iterative optimization, and then the landslide risk probability is accurately calculated. A data basis is constructed through comparison of double landslide cases, the weight calculation accuracy is improved in combination with multi-source data association analysis, key attribute description is perfected with the help of an iterative optimization mechanism, the precision and efficiency of complex mountain landslide recognition are improved, and a scientific and reliable decision basis can be provided for early warning and prevention and control of landslide disasters.

The application provides a deep energy reservoir multi-scale heat-flow-solid coupling simulation method and system, and belongs to the deep energy reservoir numerical simulation field. Spatial distribution characteristics and physical property parameters of different lithological reservoir media at a micro and mesoscopic scale are acquired, and a digital structure model of the reservoir medium is constructed. A micro and mesoscopic numerical model is obtained by embedding micro and mesoscopic action mechanisms and flow mechanisms in the model. Model structure parameters and physical property parameters under different heat-flow-solid coupling conditions are solved, and a mapping relationship is established. Based on the mapping relationship, macroscopic geological and engineering data of a target reservoir are combined to establish a macro-scale heat-flow-solid multi-field coupling control equation, and the equation is solved to output a multi-physical field response of the target reservoir under a given working condition. The application realizes dynamic transmission of physical property parameters from the micro and mesoscopic scale to the macroscopic scale and fine simulation of a multi-field coupling evolution process, and significantly improves the accuracy and reliability of deep complex reservoir engineering behavior prediction.