223 results about "Key factors" patented technology

The invention relates to the technical field of artificial intelligence, can be applied to business scenes of financial science and technology, medical health and the like, and discloses a data processing method and device based on factor weight optimization, equipment and a medium. Comprising the following steps: receiving structured data and unstructured data, and extracting a basic key factor set and monitoring object information in the data to generate a data feature set; executing multi-dimensional analysis and identification value determination based on the data feature set, and generating an analysis level and an identification value parameter; and performing weight correction processing on the basic key factor set, generating an optimized factor coefficient matrix, fusing the optimized factor coefficient matrix with associated equipment source data, and outputting a processing result. According to the method, structured and unstructured data are processed in a unified manner, so that the data integration capability is enhanced; multi-dimensional elements are extracted in combination with a pre-training model, and the key factor recognition effect is improved; and the analysis level and the identification value parameter are combined to drive weight correction and fuse the optimization matrix and the equipment source data to generate a processing result, so that the processing efficiency and the result accuracy are improved.

The invention relates to a model for evaluating and predicting mild cognitive impairment risk of old people in a pension institution. The model sequentially comprises a behavior analysis module, a language recognition module, a social modeling module, a toughness calculation module, a feature fusion module, a risk reasoning module and the like. Behavior deviation characteristics and abnormal time periods are extracted by collecting behavior data of daily life, diet, social contact and the like of old people and comparing the behavior data with an institution work and rest template; in combination with nursing records, extracting language anomaly features; analyzing social frequency and structure changes in the abnormal time period, and extracting social variation features; a cognitive toughness index is calculated by integrating the health archive and the recovery ability to the health event; and performing toughness weighting on the multi-dimensional features to construct a time sequence tensor, and inputting the time sequence tensor into a recursive model to predict a cognitive impairment risk value. And if the risk value suddenly changes, the system automatically backtracks the feature trajectory of nearly 7 days, constructs and screens a prediction path with the strongest interpretation force, outputs a dominant prediction result and a key factor sequence, and realizes high-interpretability and high-reliability early recognition and intervention reference.

The invention relates to the technical field of mechanical detection, in particular to an intelligent tool setting control method and system, and the method comprises the steps: collecting burr morphology data and tool state parameters of a slitting edge; establishing a mapping relationship between burr generation and processing parameters based on a process DOE model, wherein the process DOE model is used for extracting a key factor set influencing the slitting quality; when the real-time monitoring value of the key factor set deviates from the optimization interval of the process DOE model, generating a position and posture compensation instruction of the slitting tool rest; iteratively updating the key factor weight distribution of the process DOE model; and when the variance of the key factor weight distribution is converged to a preset threshold value, determining that the tool setting control closed-loop optimization is completed. According to the method, the defects of dynamic response hysteresis and insufficient compensation precision in traditional tool setting control are effectively overcome, and closed-loop self-optimization control of the position and posture of the slitting tool rest-process parameters-burr suppression is constructed through a three-dimensional dynamic compensation mechanism of real-time fusion of multi-source sensing data.

The invention discloses a flowmeter self-calibration method and system based on big data analysis. The method comprises the steps that multi-source data collection is conducted, and multiple kinds of data are collected and stored in a big data platform with a block chain evidence storage function; data preprocessing: adopting multiple algorithms to improve data quality; feature extraction and analysis: identifying key factors and monitoring fluid states by means of multiple technologies; calculating calibration parameters, and considering various coefficients and uncertainty analysis; performing automatic calibration, encrypting and writing parameters after digital twin simulation, and reserving a copy; and verifying the calibration effect by adopting a multi-mode verification mechanism. The system comprises a plurality of functional modules which work cooperatively. According to the invention, the data acquisition comprehensiveness is improved, the calibration period is shortened, the production efficiency is improved, the data integrity and reliability are ensured, key factors are accurately identified, the fluid state is monitored in real time, the calibration parameter adjustment is simulated, and the calibration effect is verified in a multi-mode manner, so that the reliability and accuracy of a flowmeter measurement system are improved.

The invention provides an intelligent comprehensive system and method for steel production process parameter optimization, and the system comprises a data configuration module which is used for collecting quality-related data in a steel production process; the influence factor analysis module is used for processing data related to quality by using a feature engineering technology and identifying key variables; the quality index prediction module is used for constructing a quality prediction model by using a machine learning algorithm and key variables; the process parameter optimization module is used for constructing a process parameter optimization model based on the prediction result and the quality index control target, and analyzing and recommending an optimal process parameter control range by utilizing an intelligent algorithm; and the model monitoring optimization module is used for evaluating and optimizing the operation effects of the quality prediction model and the process parameter optimization model in real time. According to the method, the whole-process intelligent management from data configuration and key factor identification to quality prediction, process parameter optimization and model dynamic monitoring is realized, a closed-loop optimization mechanism is formed, and the production efficiency and the product quality are effectively improved.

The invention provides a coal bulk loading and unloading efficiency optimization method and system based on machine learning, and relates to the technical field of coal bulk loading and unloading, and the method comprises the steps: recognizing key factors and quantitative association rules between the key factors and the coal bulk loading and unloading efficiency through a large language model; the storage yard decision topological graph is input into a reinforcement learning model, a coal bulk cargo loading and unloading scheme is determined by executing the process of state space coding, action selection and reward function calculation, and a reward function is obtained by taking a predicted value of the coal bulk cargo loading and unloading efficiency in the current period as a target and being constructed according to key factors and quantitative association rules; after the coal bulk cargo loading and unloading scheme is completed, the actual value of the coal bulk cargo loading and unloading efficiency in the current period is collected, the efficiency difference value is calculated, and the coal bulk cargo loading and unloading efficiency in the next period is optimized based on the efficiency difference value. According to the method, the efficiency can be effectively improved while various necessary conditions of loading and unloading operation are accurately met.

The invention discloses a transformer substation hidden danger point risk index model construction method, system and device and a medium, and belongs to the technical field of transformer substation risk assessment, and the method comprises the steps: collecting multi-dimensional data related to the operation of a target transformer substation; screening key factors influencing the risk of the hidden danger point, and determining the weight of each key factor; analyzing the multi-dimensional data by using a machine learning model, and identifying hidden danger points; constructing a risk assessment model and calculating risk indexes according to the weights of the key factors and the identified hidden danger points; and outputting the risk level of the hidden danger point and corresponding early warning information based on the risk index. By combining multi-dimensional data and utilizing a machine learning algorithm, the equipment operation state can be comprehensively analyzed, potential hidden dangers can be deeply mined, a large amount of historical data can be efficiently processed and analyzed, potential hidden danger points can be automatically recognized, the hidden danger point recognition accuracy is remarkably improved, early warning is ensured, equipment faults are effectively prevented, and the method is suitable for popularization and application. The operation risk of the substation is reduced.

The invention relates to the technical field of manufacturing optimization, and discloses a manufacturing process optimization decision-making method, device and equipment based on big data analysis, and the method comprises the steps: collecting data and quality evaluation information of each link of a target product manufacturing process, forming reference data, continuously recording the reference data through a big data platform, and aligning the reference data with key characteristics; according to the method, a plurality of manufacturing reference models are obtained, key factor analysis is performed on the models, optimization factors are extracted, an objective function containing quality, efficiency and cost constraints is constructed, iterative optimization solution is performed, and an optimization result is verified through a big data platform until an optimal decision scheme is obtained. The production efficiency is improved, the cost is reduced, the stability of the product quality is ensured, and the problem that in the prior art, multi-target optimization is difficult to carry out on the multi-link manufacturing process is solved.

The invention belongs to the technical field of Internet of Things, and discloses a fixed asset dynamic management method and system based on an RFID tag, and the method comprises the steps: obtaining the fixed asset data of an enterprise through employing a radio frequency identification technology, and providing the feature data of the fixed asset; based on the extracted feature data of the fixed assets, evaluating the feature data of the fixed assets by using a state analysis algorithm, and identifying key factors influencing the health state of the fixed assets; on the basis of a trend analysis algorithm, key factors influencing the health state of the fixed assets are analyzed, and the change trend of the health state of the fixed assets is obtained; and an asset maintenance plan and an operation strategy are optimized in combination with the change trend and the real-time operation state of the fixed assets. According to the invention, the real-time performance and accuracy of asset management are improved, the workload of manual checking is reduced, and the risk of loss and loss of assets is reduced, so that the overall management efficiency and decision-making quality are improved.

The invention discloses an engineering quantity list comprehensive unit price rationality detection method and system, and belongs to the technical field of engineering cost. The invention discloses a bill of quantity comprehensive unit price rationality detection method and system. The method comprises the steps of extracting key factor features; obtaining a comprehensive unit price corresponding to the project name; constructing a project recognition library; s4, correcting and perfecting the prices of the abnormal items; calculating the comprehensive unit price of the new engineering project; the system comprises a unit price calculation module, a database construction module, a comprehensive unit price identification module and a new engineering project determination module. The problem that the comprehensive unit price of the new engineering project cannot be determined in the prior art is solved, the comprehensive unit price of the new engineering project can be calculated, pricing dispute caused by fuzzy feature description of the new engineering project can be reduced, the abnormal quotation can be effectively identified, and by correcting the abnormal quotation deviating from the rationality interval, the pricing efficiency of the new engineering project is improved. The construction period delay or claim risk caused by pricing errors can be reduced.

The invention provides a comprehensive evaluation method and system for evaluating bridge capability improvement, and is applied to the technical field of data processing. The method comprises the following steps: preprocessing quantitative and qualitative indexes in basic data to generate bridge dynamic feature information; processing weight calculation related influence factor data and evaluation abnormal state information based on weight adjustment logic of an analytic hierarchy process, and generating abnormal state identification information in an evaluation process by combining dynamic factor correlation analysis of fuzzy comprehensive evaluation; processing the bridge dynamic characteristic information to generate an evaluation grade of a bridge capability improvement effect; an evaluation result, an index type and a data source in the weight calculation related influence factor data are processed in combination with a multi-stage evaluation model, and key factors influencing the bridge capability improvement effect are generated; and processing the evaluation result, the key factors influencing the bridge capability improvement effect and the abnormal state identification information, and generating dynamic adjustment information for bridge reinforcement.

The invention belongs to the technical field of business and financial integration, and discloses an intelligent business and financial integration data processing system which comprises a data management module, a factor analysis module, a risk early warning module and an optimization resource configuration module. According to the system, all the modules are integrated, fusion processing and deep analysis of business data and financial data are achieved, the system can accurately recognize key factors influencing operation and fund conditions, the perspectiveness and accuracy of risk early warning are improved, meanwhile, an enterprise is assisted in dynamically adjusting resource allocation and management and control strategies, and the system has high practicability. And the cross-department cooperation efficiency and the risk response capability are enhanced, so that fine management and robust operation of an enterprise can be realized.

The invention discloses an enhanced GAN extreme scene generation method and system considering weather factors, and relates to the technical field of scene generation in new energy power system planning and operation, and the method comprises the steps: extracting key factors from multi-time-delay meteorological data, forming an extreme label, converting the extreme label into a generation condition, and achieving the output through two-stage confrontation generation, wherein a scarce sample is stably amplified by WGAN-GP, a joint curve is generated according to conditions, and related structures, climbing and power boundary constraints are synchronously embedded to ensure statistical and physical consistency. According to the method, the authenticity of the distribution tail, the cross-variable correlation and the time structure can be considered when the sample is scarce. Extreme situation deduction of a novel power system can be supported while extreme situation coverage and credibility are improved, and an existing new energy scene analysis and planning process is integrated.

The invention relates to the technical field of traffic transportation, in particular to a railway hub passenger transport demand prediction system and method based on multi-dimensional analysis, and the method comprises the following steps: S1, multi-dimensional data collection and preprocessing: collecting multi-dimensional data related to railway hub passenger transport demands; s2, key factor analysis: analyzing the influence of key factors on the passenger transport demand in different time and space ranges; s3, regional linkage prediction: capturing a passenger flow dependency relationship of the weighted graph in a spatial dimension through a graph neural network model, and predicting a passenger transport demand of regional linkage in combination with a key factor analysis result; and S4, prediction result verification and resource optimization suggestion: generating the resource optimization suggestion for the railway hub. According to the method, specific strategies such as train number adjustment, platform scheduling optimization and peak period resource allocation are generated for railway hub operation, passenger flow fluctuation can be flexibly adapted, and resource allocation is dynamically optimized.

The invention discloses a tea processing big data analysis and quality tracing system and method, and relates to the technical field of agricultural informatization management. The tea processing big data analysis and quality tracing system is constructed, full-process data collection, real-time monitoring and accurate tracing from fresh leaves to finished products are achieved, key parameter data in the processing process are collected in real time through a sensor, key factors influencing the tea quality are extracted through an advanced data analysis algorithm, and the quality of the tea is accurately traced. According to the method, the quality of tea leaves can be accurately identified, quality abnormal points are accurately positioned, the problems of incomplete data acquisition, lagged analysis and inaccurate tracing in a traditional method are solved, meanwhile, a unique tracing code is generated for each processing batch, a full-process data mapping table is constructed, dynamic updating and real-time query of quality information are realized, the quality management level of tea leaf processing is remarkably improved, and the method is suitable for large-scale popularization and application. The requirements of consumers on transparency and high quality are met.

The present invention belongs to the technical field of power system operation and management, and provides a load-side supply and demand balancing control method and system based on multi-dimensional reliability assessment. In terms of the access mechanism, by comprehensively evaluating key factors such as the node distribution, type, capacity and dispatchability of load-side flexibility resources, scientific and reasonable evaluation standards and access thresholds are designed, which can effectively screen out high-quality and high-reliability resources to participate in the regulation, ensure that the access resources have a good regulation basis, and greatly improve the quality of the overall regulation resources; at the same time, in the centralized control mode, a global optimal control strategy is formulated to achieve efficient and unified scheduling of resources. In the distributed autonomous mode, local information perception and modeling are used, and each distributed resource optimizes local strategies based on maximizing its own interests, which can give full play to the autonomy and flexibility of distributed resources and enhance the system's ability to respond quickly to local changes.

The invention discloses a method and a system for presuming a correction coefficient of the strength of a standard test piece by adopting a small core sample test piece, and belongs to the technical field of concrete strength detection. The method solves the problems of large error and high discreteness of a presumption result caused by factors such as size effect, aggregate constraint and process difference when the standard strength is directly presumed by adopting the strength of a small core sample in the prior art, constructs a historical sample library containing multi-dimensional characteristics, calculates an initial correction coefficient and a size ratio, and obtains the presumption result of the standard strength. Key factors such as the small core sample size, the coarse aggregate particle size, the water-binder ratio, the curing age and the height-diameter ratio are comprehensively considered, an intelligent correction model is established by utilizing algorithms such as a gradient boosting decision tree and a random forest, and a comprehensive correction coefficient can be scientifically and accurately output, so that the actually measured strength of the small core sample is reliably converted into the strength of a standard test piece; and the presumption precision and the result stability are greatly improved, so that reliable technical support is provided for safety and durability evaluation of an engineering structure.

The invention provides a permafrost roadbed deformation prediction method and device and electronic equipment, and belongs to the technical field of permafrost deformation. According to the method, the influence factor data of the frozen soil roadbed can be decomposed, the data structure is simplified, dimension reduction processing is carried out on the decomposed data, and the redundancy-removed principal components are obtained through screening. Then, correlation analysis is carried out on the principal components subjected to redundancy elimination, the principal components with high correlation degree with frozen soil roadbed deformation are extracted, and high-quality input characteristics are provided for a prediction model; according to the method, the variables influencing the deformation of the frozen soil roadbed are screened, key factors with relatively high correlation with the deformation are extracted, data with relatively low correlation are removed, and the quality of a training set is improved. Therefore, on the premise of not influencing the correlation of the original training set, the data volume of the training set is reduced, and the frozen soil roadbed deformation prediction efficiency and accuracy can be improved. Furthermore, the processing advantage of the LSTM on the time sequence characteristics and the efficient calculation characteristic of the XGBoost are fused through the hybrid machine learning model, the calculation cost of frozen soil roadbed deformation prediction can be remarkably reduced while the prediction precision is ensured, and therefore the requirement for efficiency in actual engineering is better met.

The invention discloses a future pork pig group weight estimation method and system based on multi-model fusion, and the method comprises the steps: obtaining a historical structural data set of a pork pig breeding complete period, and employing a multi-modal fusion evaluation method of field priori knowledge weighted evaluation, data-driven feature importance calculation and an automatic code analysis tool. Extracting a key factor set influencing the weight of the pork pig, then based on the historical structured data set and the key factor set, respectively constructing an expert rule reasoning model, a data-driven prediction model and a code generation type dynamic model to predict the weight of the pork pig, inputting each model based on real-time data, and outputting a prediction value; weights are dynamically distributed according to historical errors, results are fused, abnormal values are removed in combination with discrete degree detection, and a final fusion prediction value is generated. The technical problem of improving the weight estimation accuracy of the pork pig group in the prior art is solved.

The invention relates to the technical field of building anti-seismic risk assessment, and discloses a building anti-seismic potential safety hazard assessment method and system based on multi-factor coupling, and the method comprises the following steps: collecting key factors of potential safety hazards of a to-be-assessed building, and endowing each piece of key factor information with a discrete value; constructing a nonlinear coupling model through key factor construction and an interaction effect between key factors; building state information is monitored in real time through a sensor, a damage or collapse possibility index is corrected according to monitoring data to obtain a correction value, and a second comprehensive risk value is obtained through recalculation; and according to the second comprehensive risk value obtained through calculation, building anti-seismic safety risk grades are divided. According to the method, the damage possibility index and the accident consequence severity index are constructed, the comprehensive risk value is calculated according to the damage possibility index and the accident consequence severity index, quantitative modeling evaluation from static attributes to the risk value is achieved, and the method is more scientific and objective compared with a traditional single-factor evaluation method with high empirical and subjective performance.

The invention relates to the technical field of asset assessment. The asset evaluation intelligent system based on the AI technology is provided, and the system comprises the steps that web crawler capture processing is conducted on a data source of a public network channel, and a first type of heterogeneous data is obtained; aPI interface docking processing is carried out on the data source of the enterprise internal system, and second-class heterogeneous data is obtained; performing logic splicing processing on the first type of heterogeneous data and the second type of heterogeneous data to generate multi-source heterogeneous asset evaluation data; performing model matching processing on the asset value characteristic matrix based on the asset type, and selecting an estimation model architecture; performing interpretability analysis processing on the dynamic evaluation parameters to generate a key factor contribution degree and an interpretable evaluation conclusion; and performing document processing on the contribution degree of the key factor and the interpretable evaluation conclusion to generate an interactive evaluation document so as to improve the multi-source data integration capability, enhance the risk dynamic evaluation adaptability and strengthen the real-time relevance and interpretability of the evaluation document.

The invention discloses a method and system for constructing power operation personal risk dynamic evolution through key factor analysis, and relates to the field of artificial intelligence, and the method comprises the steps: constructing an operation risk analysis knowledge graph, and carrying out the entity recognition; extracting knowledge graph entities; performing accident chain replay of the knowledge graph and the expert model; and constructing a risk dynamic evolution model based on the space-time knowledge graph. According to the method, previous electric power operation accident cases are analyzed, influences of different time and different environments on operation of electric power personnel are researched, risk factors in historical accidents are analyzed, an analysis method based on a space-time knowledge graph is provided, historical accident chain deduction is achieved, and the risk factors in the historical accidents are analyzed. And providing a power operation personal risk dynamic evolution model based on key risk factors.

The invention discloses a tree safety risk assessment method, system and device based on a key index priority mechanism, and a storage medium. The method comprises the following steps: constructing an inference model according to a fault tree model of a street tree and a fuzzy Bayesian network; based on the inference model, through reverse inference, sensitivity analysis and most approximate cause chain analysis, screening out a key index set and a conventional index set from a preset risk candidate index set; and based on a hierarchical risk assessment principle, respectively assessing the key index set and the conventional index set to obtain an assessment result. According to the inference model, key factors having significant contributions to the overall risk of the border tree are identified from a large number of risk indexes; a conjoint analysis method of reverse reasoning, sensitivity analysis and most approximate cause chain analysis is adopted, so that false alarm and missing alarm are effectively reduced; according to the hierarchical risk assessment principle, key factors are assessed preferentially, it is ensured that the street tree with structural fatal defects can be recognized and controlled in time, and the reliability and timeliness of assessment are improved.

The invention discloses an industrial robot automatic control system and method, and relates to the technical field of industrial robots, and the system comprises a data management module, a data analysis module, a motion prediction module and a coordination control module. By accurately analyzing key factors, the system can identify core parameters influencing the motion performance of the robot, the adaptability and working efficiency of the robot in a complex environment are optimized, the motion precision and stability are improved, the system predicts the future motion state and track of the robot in a prospective mode by means of a motion prediction model, and the robot motion performance is improved. Potential fault risks are recognized in advance, production interruption is avoided, the service life of the robot is prolonged, motion parameters are dynamically adjusted according to a prediction result, a control strategy is optimized, and the flexibility and execution efficiency of the robot in a multi-task scene are improved, so that the production efficiency and the product quality are remarkably improved, and the labor cost and operation loss are reduced. And greater economic benefits are created for enterprises.

The invention discloses a power distribution network fault self-healing success rate prediction and strategy evaluation method, system and device and a medium, and relates to the technical field of power distribution network self-healing, and the method comprises the steps: collecting power distribution network operation data and historical self-healing case data, analyzing key factors influencing the self-healing strategy success rate after preprocessing, and constructing an influence factor network; a probability graph model is established based on the network to perform success rate prediction, a classification evaluation model is constructed to perform preliminary evaluation, a prediction result is output through multi-model integration optimization fusion, a strategy evaluation index system is constructed to perform comprehensive evaluation on a self-healing strategy, and an adaptive evaluation mechanism is introduced. Dynamically adjusting model parameters according to the network state similarity, and finally outputting an evaluation result. The power distribution network fault self-healing success rate prediction and strategy evaluation method realizes the technical spanning of power distribution network fault self-healing success rate prediction and strategy evaluation from single dimension to multiple dimensions, from static analysis to dynamic optimization, and from simple prediction to comprehensive evaluation, and significantly improves the intelligent level and decision reliability of power distribution network self-healing operation.

The invention relates to the technical field of engineering slope instability risk analysis, in particular to a weak intercalated layer-containing bedding slope instability risk analysis method, equipment and a medium. According to the technical scheme of the invention, a complex and uncertain slope engineering problem is converted into a computable and reasonable probability graph model. Full life cycle disaster control factors from natural conditions, geological conditions, engineering design, construction and operation maintenance are systematically combed, omission of key factors is avoided, and comprehensiveness of risk analysis is guaranteed. The traditional qualitative causal analysis is upgraded to a quantitative probability dependency relationship, and a mathematical model basis is provided for subsequent risk diagnosis and prediction. And the crossing of risk analysis from'static evaluation 'to'dynamic reasoning' is realized. The model not only can give a static risk value, but also can dynamically update the probability of all related variables after new evidence (such as monitoring data) is obtained, so that intelligent diagnosis and prediction are realized.

The invention belongs to the technical field of supply chain management, and particularly provides a supply chain management method based on multi-source data, and the method comprises the steps: obtaining the sales data of a retailer in a preset time, and carrying out the preprocessing; analyzing the preprocessed sales data, marking consumers associated with the key factors, and evaluating the influence weights of the key factors on the purchase behaviors of the marked consumers; acquiring and calculating historical sales data and all key factor data within a preset time to obtain enhanced prediction features; constructing a demand prediction model, inputting the preprocessed sales data, the enhanced prediction features, the marked consumers and the evaluated influence weight into the demand prediction model by using a learning algorithm, predicting the supply demand of each node of the supply chain within a preset time through the demand prediction model, and generating a corresponding supply strategy; the supply strategy of the supply chain is flexibly adjusted by integrating multiple data sources, so that quick response to market change and efficient management of the supply chain are realized.

The invention relates to the technical field of environmental risk management and control, in particular to a tire wear particle environment fate prediction model construction method based on multi-dimensional data. Comprising the following steps: data acquisition and preprocessing; a feature importance analysis model based on a random forest is constructed, key factor mining is performed on the preprocessed collected data, core factors influencing tire wear particle environment fate are screened by calculating Gini importance values of the key factors, and feature vectors and confidence coefficients of the core factors are output; classifying and integrating factors; carrying out model association construction; model training storage; and performing model verification feedback. According to the method, the continuity, timeliness and consistency of tire wear particle sample data and environment associated data are improved through the built tire wear particle environment fate prediction model and through cooperative collection, space-time registration and noise filtering preprocessing of the multi-source monitoring equipment, and a high-quality data basis is provided for subsequent model building.

The invention provides a scientific research project effectiveness evaluation method based on an LSTM network, and the method comprises the steps: firstly constructing a scientific research project effectiveness evaluation index system which covers key factors such as project innovation, scientific research output, fund use efficiency and social influence; then, historical data, including index data, historical weights and evaluation scores after project implementation, of previous scientific research projects are collected, and training is carried out based on the LSTM network; and for a to-be-evaluated scientific research project, obtaining validity evaluation index data of the to-be-evaluated scientific research project and an index weight of a previous scientific research project, inputting the data into the trained LSTM network, predicting the weight of each index, and calculating an evaluation score in combination with the weights. The loss function comprehensively considers a prediction evaluation error, a prediction weight error and a next time node weight influence error, and ensures that a prediction result is more accurate. Scientific and dynamic evaluation can be provided before implementation of a scientific research project, decision reasonability is improved, and powerful support is provided for scientific research resource allocation and management.

The invention provides a method and a device for predicting the single-layer productivity of a multi-layer gas reservoir. Obtaining target features of a target layer related to the target well; wherein the target features comprise geological features and fracturing features; obtaining an initial prediction capacity corresponding to the target layer based on the target features of the target layer by using a preset capacity prediction model; wherein the preset productivity model is obtained by combining contribution degree analysis and feature weight adjustment in advance and utilizing sample data training; and according to the initial prediction capacity corresponding to the target layer, an exploitation strategy for the target layer of the target well is determined. Based on the method, the geological features and the fracturing features of the target layer are deeply analyzed by presetting the productivity prediction model and combining contribution degree analysis with feature weight adjustment, so that fine decoupling between the initial prediction productivity and each key factor is realized, the precision and the interpretation ability of single-layer gas production prediction are improved, and the prediction efficiency is improved. And powerful support is provided for formulating a more accurate and scientific mining strategy.