28 results about "Local evaluation" patented technology

The invention provides a multi-layer satellite network routing method and system based on multi-agent deep reinforcement learning. The method comprises the following steps that: a plurality of low-orbit satellites train respective executor networks by adopting a distributed multi-agent reinforcement learning algorithm to realize routing decision; and one medium-orbit satellite trains a global evaluation network according to the local parameters of the performer network uploaded by the plurality of low-orbit satellites, and provides the global parameters of the trained global evaluation network to each low-orbit satellite, so that the low-orbit satellites update respective local evaluation networks according to the global parameters. And updating respective executor networks according to the updated local evaluation network. According to the multi-layer satellite network routing method and system based on multi-agent deep reinforcement learning, dynamic optimization of a routing strategy can be realized in a medium and low orbit satellite network, and the overall performance of the satellite network is remarkably improved.

The invention discloses a method for stopping a spacecraft target access optimization task based on local search and exploration enhancement, and belongs to the technical field of spacecraft control. In order to solve the problem of optimization termination control with robustness and self-adaptability, an optimization termination criterion is designed to terminate iterative optimization of Bayesian optimization when the maximum prediction standard deviation of a Gaussian process regression agent model in a current local area is lower than a preset threshold value; the design optimization termination criterion is embedded into the BO; measuring the cognitive uncertainty of the GPR agent model to the target function at a certain point by using the prediction standard deviation of the GPR agent model; and setting a judgment condition for judging whether to enhance the exploration enhancement capability of the acquisition function and whether to trigger local evaluation. According to the invention, optimization termination control with robustness and self-adaptability is realized.

The invention discloses a concrete column axial deformation identification method and system based on a visual image, and relates to the technical field of building structure monitoring, and the method comprises the steps: collecting a continuous image sequence of a concrete column from different sides; edge detection and morphological processing are carried out on the continuous image sequence, the overall contour of the cylinder is extracted, the axial center line of the cylinder is fitted, and global axial length change features are generated; executing adjacent frame tracking identification of the local evaluation center point, and establishing a first local deformation feature; performing local deformation feature extraction of the convolutional neural network, and establishing a second local deformation feature; and hierarchical fusion is carried out, and an axial deformation identification result is established. According to the invention, the technical problems of one-sided deformation information, insufficient monitoring precision and poor reliability caused by single feature extraction of the axial deformation of the concrete column in the prior art are solved, and the technical effects of realizing multi-source feature fusion identification of the axial deformation of the concrete column and improving the precision and reliability of deformation monitoring are achieved.

The invention provides a copper-aluminum transition splicing sleeve performance evaluation method, system and equipment under a long-term working condition, and the method comprises the steps: carrying out the material performance test of a to-be-tested pipe, and obtaining a material performance evaluation result; placing the plurality of to-be-tested tubes in a multi-loop parallel large-current test structure for synchronous electrical contact performance test, and performing temperature bidirectional compensation to obtain an electrical contact performance evaluation result; performing a stability simulation test on the to-be-tested pipe to obtain a long-term operation reliability evaluation result; synthesizing the three evaluation results of the to-be-tested pipe to obtain a comprehensive evaluation result of the to-be-tested pipe; according to the method and system, overall performance evaluation is achieved by comprehensively considering the aspects of material performance, electrical contact performance and long-term operation reliability, then temperature bidirectional compensation is carried out on the electrical contact performance of the multi-loop parallel large-current test structure, so that the precision of an electrical contact performance evaluation result is improved, local evaluation is achieved, and the reliability of the test structure is improved. And a set of complete high-precision copper-aluminum transition splicing sleeve performance evaluation method is formed.

The invention introduces an Internet of Vehicles vehicle reputation dynamic evaluation method based on an evaluation graph. The method comprises the following specific steps: S1, constructing a global vehicle evaluation graph; s2, carrying out local evaluation and set division; s3, evaluation set selection; s4, performing weighted aggregation calculation; s5, performing iterative updating and amplitude limiting processing; s6, vehicle-trust center closed-loop interaction is carried out; s7, performing long-term reputation evolution; according to the method, evaluation control and collusion attacks are prevented, long-term and smooth evolution of the vehicle reputation is also realized, the robustness and credibility of an Internet of Vehicles system are remarkably improved, meanwhile, forward excitation on low-reputation vehicles is considered, and a dynamic and credible vehicle reputation management system is formed.

A local-global template search method of multi-domain protein, first, input the three-dimensional structure of each single domain of the multi-domain protein to be assembled, set the number R of templates to be screened in local evaluation; then, score the alignment of each template in the multi-domain protein structure library with each two consecutive domains, respectively calculate the local score of each two consecutive domains from N terminal to C terminal and from C terminal to N terminal, select the template with the highest local score as the optimal local template; then, again for each template in the multi-domain protein structure library, perform local evaluation by structure alignment of each domain and template through TM-align, and calculate the score of the template according to the local score of each single domain, select the top R templates with high score for global evaluation; in global evaluation, the alignment of different domains is not allowed to overlap, select the template with the highest global score as the optimal global template; finally, output the optimal local template and the final full-length template. The application improves the search precision of the template.

The invention provides a high-speed railway joint debugging and joint test evaluation method and device, and the method comprises the steps: determining a high-speed railway joint debugging and joint test detection index according to the operation data of a high-speed railway system; cloud processing is carried out on the detection indexes to obtain cloud feature parameters of the detection indexes; calculating the information entropy of each detection index based on an entropy weight method; determining the weight of each detection index according to the information entropy of each detection index; according to the cloud feature parameters of the detection indexes and the weight of each detection index, a high-speed railway joint debugging and joint test comprehensive evaluation cloud model is obtained through a weighted average method; according to the high-speed railway joint debugging and joint test comprehensive evaluation cloud model, determining a high-speed railway joint debugging and joint test evaluation result; according to the method, the transformation of the high-speed railway joint debugging and test from local evaluation to system overall comprehensive evaluation is realized, the fuzziness and randomness in the evaluation are effectively quantified through the combination of the cloud feature parameters and the weights, the reliability of the evaluation result is improved, and a data reference is provided for railway engineering acceptance.

The present invention discloses a method for evaluating the degree of fusion of regenerated asphalt mortar, comprising the following steps: S1, obtaining force-displacement curves of asphalt mortar calibration specimens with different old asphalt mixing ratios, and calculating the interface failure energy of the corresponding calibration specimens based on the force-displacement curves; S2, fitting a regression equation for the interface failure energy and the degree of fusion; S3, obtaining the force-displacement curves of the regenerated asphalt specimens to be evaluated and calculating the interface failure energy, and calculating the degree of fusion based on the fitted regression equation. The present invention determines the quantitative evaluation of the degree of fusion through a pull-out test, which can provide a basic database for reference of process parameters for the preparation of regenerated asphalt mortar. The interface failure energy is easy to obtain, and the evaluation method can be reused for the same raw material. Moreover, the fusion effect of the entire fine material in the mortar can be evaluated rather than a local evaluation.

The present disclosure provides a server parameter optimization method and device, wherein the method comprises: obtaining a state parameter set and an execution action set of a server cluster, the state parameter set and the execution action set having a time label; based on the state parameter set and the execution action set, determining a local evaluation prediction parameter by using a multi-agent reinforcement learning algorithm; based on the state parameter set and the execution action set, determining a global evaluation prediction parameter by using an adaptive dynamic programming model; and based on the local evaluation prediction parameter and the global evaluation prediction parameter, optimizing the execution action set. The method of the present disclosure determines the local evaluation prediction parameter and the global evaluation prediction parameter corresponding to the server cluster by using the QMIX model and the ADP model, thereby realizing the automatic optimization of the execution action set of the server cluster, so as to improve the stability and service efficiency of the server cluster.

A kind of stop method of space vehicle target access optimization task based on local search and exploration enhancement belongs to space vehicle control technical field.To solve the problem of optimization termination control with robustness and self-adaptation, the present application designs optimization termination criterion as when the maximum prediction standard deviation of Gaussian process regression surrogate model in current local area is lower than preset threshold, terminate the iteration optimization of bayesian optimization;The design optimization termination criterion is embedded into BO;The prediction standard deviation of GPR surrogate model is used to measure the cognitive uncertainty of GPR surrogate model to objective function at a point;Set the judgment condition to judge whether the exploration enhancement ability of acquisition function is enhanced and whether local evaluation is triggered.The present application realizes the optimization termination control with robustness and self-adaptation.

The invention discloses a method and system for verifying and evaluating the design of human factors engineering (HFE) for a nuclear power plant, and relates to the technical field of verification and evaluation of human factors engineering. The invention is based on HFE influencing factors. On the one hand, a subjective evaluation data table is designed for each influencing factor. On the other hand, local and global evaluation methods are constructed under different categories of influencing factors to perform local and global multi-objective evaluations, thereby obtaining local evaluation results under each category indicator system and global evaluation results of all indicators under the category. According to the local or global evaluation results, the input values ​​of the HFE factor levels in the evaluation process can be traced back to discover the defects of one or some indicators and the overall design level, and timely and effective feedback, interface adjustment and optimization can be performed.

The invention relates to the technical field of artificial intelligence security and natural language processing, and particularly provides a model security availability joint evaluation method based on dynamic multi-objective optimization. The method comprises the following steps: iteratively updating a test suffix by using an optimization signal provided by a local reference model to generate an evaluation sample; a local evaluation sample optimized based on a reference model is applied to a to-be-tested target model environment for verification and evaluation, and the method solves the problems that evaluation target coverage is insufficient, and sample naturalness and scene consistency are weak.

The invention discloses a method for evaluating urban space boundary recognition accuracy, a medium and a program product. The method comprises the steps of obtaining a coordinate point set created based on a research area range; obtaining a reference set and a prediction set; a preset evaluation step is repeatedly executed, an overall analysis accuracy evaluation result is obtained, and the evaluation step comprises the steps that the intersection-to-union ratio is calculated based on the reference set and the prediction set, and then the matching relation of polygons in the reference set and the prediction set is judged; for the uniquely matched polygon pair, calculating a Hausdorff distance and a turning function distance; the number of the polygon under the non-unique matching condition and the number of the polygon under the unique matching condition with the intersection-to-union ratio smaller than a set threshold value are fed back to the analytical model, so that the analytical model is locally optimized; and respectively marking polygons in the reference set and the prediction set in the coordinate point set to obtain a first list and a second list, and carrying out confusion matrix analysis to obtain an overall analysis accuracy evaluation result. According to the invention, overall and local evaluation can be carried out on urban space boundary analysis accuracy.

The invention relates to the technical field of platform management, and particularly discloses an enterprise AI intelligent platform management method based on edge computing, which comprises the following steps of: performing distributed processing on data divided according to service types on an edge side, extracting key service indexes by each edge node, calculating an index state score by comparing a current value with a dynamic statistical reference value generated based on historical data, and aggregating to generate a local evaluation result; on the center side, the central management platform gathers all edge data, constructs an enterprise operation association map reflecting complex association among all business indexes, simulates the influence of local fluctuation on the overall situation through map calculation based on the map, and finally generates an enterprise comprehensive score comprehensively considering the overall performance, the critical path health degree and the system toughness. Therefore, closed-loop management from local intelligence to global optimization is realized.

A hand-held device, such as a remote control, is provided for operation of a consumer appliance and is equipped with one or more biometric sensors to capture and report on the health condition of a user of the hand-held device. The data captured by such a hand-held device may be evaluated locally in the hand-held device itself or may be conveyed to a target consumer appliance either for local evaluation by that consumer appliance or for onward transmission to a central off-site monitoring service.

The invention discloses a multi-agent collaborative reinforcement learning method, device and equipment, a medium and a product. The method comprises the following steps: constructing an experience pool according to historical action parameters of multiple agents, selecting target experience data from the experience pool, and determining next-step execution strategy parameters of the multiple agents according to the target experience data and a multi-agent collaborative network; global return data and local return data are determined according to the next execution strategy parameters of the multiple agents; and performing parameter updating on the multi-agent collaborative network based on the global evaluation network and the local evaluation network according to the action parameters, the strategy parameters, the global return data and the local return data of the multiple agents. According to the method, the optimization efficiency of the intelligent agent strategy is maximized through the double-evaluation network architecture, the global evaluation network is responsible for capturing the macroscopic situation of multi-agent interaction, the local evaluation network focuses on strategy optimization of an individual behavior level, accurate updating of the strategy is achieved through the synergistic effect of the global evaluation network and the local evaluation network, and then the cooperation ability of the intelligent agents is improved.

The application provides an enterprise security comprehensive analysis method and system based on multi-modal data, and relates to the technical field of distributed processing. The method comprises the following steps: collecting local security data of each work area edge evaluation node and forming multi-modal security data, combining node resource constraint portraits to perform heterogeneous load distribution and processing depth self-adaptive adjustment to obtain standardized multi-modal security data segments, further generating a local security representation vector through feature space mapping and constructing a local risk relationship graph, calculating node evaluation confidence parameters and node drift correction parameters on the basis to complete confidence weighting and deviation correction of the local evaluation result, realizing semantic alignment and topological splicing of multi-node results by using a cross-node alignment mapping matrix to generate a global security evaluation graph, finally performing risk aggregation determination and iterative correction combined with dynamic updating parameters, and outputting a global security evaluation result.

The invention relates to the field of network security, and particularly discloses a multi-agent cooperative defense and evaluation method, which comprises the following steps of: independently acquiring original data in a monitoring range and generating local evaluation evidences through a plurality of agents deployed at different nodes; in order to solve the problem that evidence sources are different in credibility, an honor sub-mechanism is introduced to carry out weighted discount on local evaluation evidences. Meanwhile, a selective broadcasting strategy is designed, and broadcasting is carried out only when evidences meet the importance or variability threshold value. After all weighted and screened evidences are collected, heterogeneous and asynchronous evaluation information from different local perspectives is aggregated into a unified and reliable global consensus result through global evidence fusion and a consensus decision mechanism. And finally, the defense coordinator generates and issues a defense instruction according to the consensus result, so that the technical problem that accurate global threat cognition is difficult to quickly form due to view angle limitation, model heterogeneity and communication overhead in the traditional method is effectively solved.

The present invention discloses a remote sensing image mosaicking method based on image quality, which includes the following steps: obtaining evaluation indicators and establishing first, second, and third evaluation criteria respectively according to the evaluation indicators through weighted summation operations; screening images to be mosaicked according to the first and second evaluation standards and determining the mosaicking order to obtain a mosaicking scheme, and performing virtual mosaicking according to the mosaicking scheme to obtain a virtual synthetic image; performing an overall evaluation of multiple virtual synthetic images based on the third evaluation standard to select the optimal mosaicking scheme; and mosaicking the images to be mosaicked using image mosaicking technology according to the optimal mosaicking scheme. This method not only performs local evaluation on the individual images involved in the mosaicking, but also performs an overall evaluation on the virtual mosaicked synthetic image after virtual mosaicking, thereby ensuring that the quality of the images involved in the mosaicking is optimal and that the overall quality of the mosaicked synthetic image is at a high level.

The embodiment of the application provides a kind of based on Gaussian function's plate shape evaluation method, device, equipment and storage medium, relate to metallurgical rolling field, by using flatness tester to the preset sampling point on the steel plate to be detected, obtain the height data measured by flatness tester, according to the height data, construct plate shape matrix, using Gaussian function to plate shape matrix is fitted, obtain plate shape defect information, construct local evaluation function and overall evaluation function, according to the plate shape defect information, local evaluation function and overall evaluation function, the steel plate to be detected is evaluated, according to the evaluation result, determine whether the steel plate to be detected is qualified, according to flatness tester data, accurately obtain the defect information of steel plate plate shape, according to defect information, accurately evaluate the qualification of steel plate, improve the steel plate classification precision, save manpower and material resources, improve the production efficiency of steel plate.

The invention relates to the technical field of cultural tourism resource evaluation. The multi-dimensional cultural tourism resource comprehensive evaluation method and device, the equipment and the medium are provided, and the method comprises the steps that an initial local evaluation model is jointly trained through a preset federated multi-task framework, and a shared base layer and a personalized task layer are obtained; aggregating parameters of the shared base layer to generate a global consensus model; combining the personalized task layer and the global consensus model to generate a terminal exclusive evaluation model; anonymization weight vectors in the terminal exclusive evaluation model are extracted, and an anonymization weight vector set is obtained; analyzing conflict dimensions of the anonymized weight vector set, and generating a multi-dimensional value conflict graph; and performing group decision negotiation based on the terminal exclusive evaluation model and the multi-dimensional value conflict graph, and outputting a Pareto improvement strategy list, so as to achieve the technical effects of improving data privacy protection intensity, balancing global consensus and local characteristics, and optimizing conflict decision quality.

An evaluation method and device for the load prediction results of a neural network based on SHAP values. The method includes: extracting the load prediction data at each moment output by the neural network load prediction model integrating the attention mechanism; the extracted prediction data passes through the calculation module of the load influence feature contribution degree to obtain the load influence feature contribution degree; inputting the load influence feature contribution degree into the evaluation module for evaluation to obtain an evaluation result; the evaluation module includes a global evaluation module and a local evaluation module, the global evaluation module is used to evaluate the magnitude of the influence of each feature on the load and give the importance ranking of the influencing factors; the local evaluation module is used to dynamically evaluate the contribution degree of the input features and the contribution degree of the time series features at each moment, and further evaluate the reasons for the generation of the prediction results. The present invention can evaluate the prediction results of a model with the "black box" nature of the neural network, analyze the optimization effect of the attention mechanism on the load prediction model, and improve the reliability of the load prediction model.

The invention provides a multi-layer satellite network routing method and system based on deep reinforcement learning and federated learning, and the method comprises the following steps: a plurality of low-orbit satellites train respective executor networks through employing a distributed deep reinforcement learning algorithm, so as to achieve a routing decision, training respective evaluation networks to achieve local evaluation of the performer networks; and one medium-orbit satellite generates a global evaluation network based on a federated learning algorithm according to the evaluation network parameters of the evaluation network uploaded by the plurality of low-orbit satellites, and provides the global evaluation parameters of the trained global evaluation network to each low-orbit satellite, and sending the global evaluation parameters to the low-orbit satellites, so that the low-orbit satellites update respective evaluation networks according to the global evaluation parameters, and update respective executor networks according to the updated evaluation networks. According to the multi-layer satellite network routing method and system based on deep reinforcement learning and federated learning, dynamic optimization of satellite routing is realized based on the federated learning technology and the deep reinforcement learning technology.

The invention provides an environmental protection big data supervision platform based on a block chain. The invention discloses an environmental protection big data supervision platform based on a block chain, and the platform comprises a monitoring module which is used for carrying out the atmospheric environment monitoring of a target region, and obtaining environment monitoring data; the data storage module is used for storing the environment monitoring data in a target block chain node; the data analysis module is used for evaluating and analyzing the environment monitoring data to obtain an analysis result; and the local evaluation module is used for calculating an evaluation value of a local area in the target area according to the atmospheric environment score and a preset score threshold. The atmospheric environment pollution degree of the target area is locally analyzed through the data analysis module, some special areas in the target area are further analyzed through the local evaluation module, deep mining of the monitored atmospheric environment data is achieved, and the utilization rate of the data is improved.

The invention discloses a multi-dimensional contribution degree-based knowledge sharing excitation method and device and a storage medium, and relates to the technical field related to the Internet, and the method comprises the following steps: constructing and distributing cross-scene adaptation state metadata; the local node executes local evaluation of the knowledge contribution degree and generates feedback data based on the metadata and the real-time demand data of the local scene, and writes the feedback data into a local adaptation result of the metadata; the aggregation node collects metadata of a plurality of local nodes, and performs federal aggregation on local evaluation related parameters in the metadata; performing full-link adaptation effect simulation verification of knowledge cross-scene circulation based on the updated global optimization parameters, and optimizing metadata according to a verification result; the optimized metadata are distributed to a local node and an aggregation node, the local node updates a knowledge contribution degree evaluation rule, and the aggregation node calibrates local model parameters used for parameter aggregation; the problem of mismatching of excitation resources is solved.

This invention discloses a knowledge graph-based method for evaluating human clinical information, comprising the following steps: collecting human clinical information of a target individual to generate a standardized clinical dataset; performing medical semantic extraction to generate a set of medical semantic elements; constructing a patient-level evaluation knowledge graph for the target individual to generate a clinical efficacy structure; identifying risk endpoint nodes, constructing an Even-Shiloach tree to form an Even-Shiloach forest; identifying a set of invalid edges, deleting the edges corresponding to the invalid edge set to obtain an updated Even-Shiloach forest; generating a hierarchical difference set and identifying the affected subgraphs, performing local evaluation updates to generate corresponding evidence chains; determining the initial evaluation result of the human clinical information, correcting the initial evaluation result to obtain a corrected evaluation result. This invention achieves dynamic updability and traceable evidence chains in the evaluation of human clinical information.

An object of a system according to an embodiment is to appropriately evaluate thinking skills and subjective learning.SOLUTION: A system includes an analysis part, an evaluation part, and an indication part. The analysis unit analyzes a sentence of a student. The evaluation unit evaluates thinking ability based on the sentence analyzed by the analysis unit. The pointing unit points out an error or an improvement point based on a result evaluated by the evaluation unit.SELECTED DRAWING: Figure 1

The present application relates to the technical field of deep geothermal prediction and detection, and particularly relates to a multi-element fusion deep geothermal detection method, comprising the following steps: constructing a multi-element evaluation structure and mapping it into a score propagation network, obtaining multi-source basic data to generate element scores and uncertainty information, extracting time changes to construct time elements, introducing a nonlinear coupling relationship to dynamically correct the score propagation, triggering constraint rules based on uncertainty to regulate the propagation amplitude and path, recording the influence path and contribution ratio through score propagation, and finally forming a comprehensive score corresponding to the target area comprehensive node, realizing the structured, dynamic and interpretable evaluation of the deep geothermal system. In the present application, by constructing a multi-element evaluation structure and mapping it into a score propagation network, the system integration and step-by-step transmission of multi-element information are realized, the one-sidedness of traditional single-parameter local evaluation is improved, and the comprehensive reflection ability of deep geothermal characteristics is improved.