4837 results about "Test data" patented technology

Systems and methods disclosed herein automatically evaluate, select, and coordinate artificial intelligence (AI)-based agents for collaborative distributed task execution based on dynamic, multi-attribute scoring and resource allocation models. The system obtains a task specification request defining a computational requirement set, a performance metric set, and an available resource set for one or more tasks to be executed by a network of AI-based agents. A first AI model set generates domain-specific test datasets and validates prospective agents by comparing agent-generated fingerprints against predetermined hash values stored on a distributed or federated ledger. A second AI model set constructs a multi-dimensional scoring data structure for each agent by using historical performance metrics to compute weighted composite scores. The system selects a subset of AI-based agents, ranks the agents, and allocates resources proportional to each agent's composite score. A third AI model set coordinates and executes distributed computer-executable workflows across the selected agents.

The present invention relates to the technical field of deep learning and computer vision. Disclosed is an unmanned aerial vehicle image-based small object detection method for target areas. The present invention crops images of obvious small objects in certain target areas, and annotates the small objects of different categories to form a raw training and testing dataset, so as to ensure the accuracy of data required in the early stage of the algorithm and further ensure the scientificity of the algorithm; uses the computing capability of an improved YOLOv7 detection model to collect image features of different degrees in the dataset, the improved YOLOv7 detection model using YOLOv7 as a basic model and adding to a neck network an MS-CET module, which is constituted by an improved self-attention mechanism and convolution module SPPCSP, and a BHC-FB module, which is constituted by bidirectional mixed convolution modules NConv and RPConv connected in parallel; and finally fuses different feature layers as a final judgment basis of an unmanned aerial vehicle for small object detection in the target areas, to further check the accuracy of the algorithm and criteria for dataset selection, thereby improving recognition accuracy.

The invention relates to the field of electric power material quality detection, and discloses a test scheduling system for detection task collaboration and data acquisition, which comprises a task construction module, a state collaboration module, a graph reasoning module and a data acquisition module. And the task construction module generates a standardized test task packet including a task identifier, a project code, a target equipment identifier, an environment requirement parameter and a two-dimensional code according to the test rule base and the resource configuration state, and pushes the standardized test task packet to corresponding test equipment through a Web Service interface. And the state collaboration module receives an equipment state feedback event, constructs an event time sequence flow graph based on the task identifier and generates a task state sequence with a timestamp. The atlas reasoning module takes the state sequence and the environmental parameters as input, constructs a test atlas structure and generates an optimization execution path. And the data acquisition module controls the test equipment to complete a detection task according to the path, acquires test data and environmental parameters, and encapsulates the test data and the environmental parameters to form a structured task data packet, thereby realizing data collection and task tracing.

The invention discloses a railway vehicle test data management and analysis system, and the system comprises a multi-modal data collection module which collects the heterogeneous data of tests such as airtightness and weighing in real time; the block chain credible evidence storage module is used for performing time-space stamp marking and hash encryption on the data and verifying the integrity; the space-time atlas analysis module is used for constructing a space-time heterogeneous atlas and generating vehicle-level feature vectors through graph convolutional network fusion data; the predictive maintenance decision module is used for predicting subsystem performance degradation and fault risks and generating a hierarchical maintenance strategy; and the adaptive visual platform integrates a large screen and a mobile terminal to display data and decisions. According to the system, efficient integration and credible evidence storage of multi-source data are realized, the data utilization efficiency and analysis depth are improved, accurate maintenance decision is supported, and safe operation of railway vehicles is guaranteed.

The invention relates to the technical field of hydrogen energy automobiles, in particular to a method for establishing a hydrogen leakage prediction model of a hydrogen energy automobile. According to the method, training, verification and test data sets are obtained through multi-modal data fusion and screening based on physical significance on source data fusing CFD simulation data and actually measured leakage data, multi-dimensional information is extracted from training data, time domain, frequency domain and physical field feature matrixes are obtained, a time sequence coding layer and a physical decision layer are constructed on the basis, and the time sequence coding layer and the physical decision layer are constructed. Therefore, fluid mechanics is integrated into a model structure, the model can explain and make decisions according to physical laws, the problem that fault attribution analysis conforming to the physical laws cannot be provided in the prior art is solved, a hybrid architecture model integrating data driving and physical constraint is constructed based on a two-layer structure, data set optimization is verified, and a test data set is adjusted. And finally, a hydrogen leakage monitoring model with high-precision prediction capability and reliable physical interpretability is established.

The invention discloses a software test automation report generation optimization method which comprises the following steps: automatically registering and authenticating a multi-source test data access endpoint, uniformly collecting and normalizing test data with a multi-dimensional attribute tag, and forming a high-quality grouped data set through structured modeling and automatic clustering; further extracting event and result feature nodes, and establishing an interpretable tree rule dependency relationship; according to the method, the minimum redundancy and reusable customized rule path is generated by applying a dependency tree traversal and optimization algorithm and automatic reasoning, and the continuous adaptive evolution of the rule base among multiple projects is realized by combining conflict detection and visual interaction, so that the efficiency and expandability of data processing in a complex test scene are improved, and the test efficiency is improved. And the flexibility and automation level of report customization rule configuration are enhanced.

The invention relates to a test case generation method and system based on an industrial agent. The method comprises the steps that an industrial data set is acquired, and a system dependency model is constructed; generating a path coverage test sequence through the first agent; generating a boundary value test data set through the second agent; generating abnormal test scene information through a third agent; and generating a test case set based on the path coverage test sequence, the boundary value test data set and the abnormal test scene information. In conclusion, by constructing the system dependency model and utilizing the multi-agent collaborative analysis process logic path, boundary condition and dependency relationship, the multi-surface coverage of the industrial control system test scene is realized, the automation of the test case generation is realized, the effects of improving the generation efficiency and enhancing the comprehensiveness of the test coverage are achieved, and the method is suitable for industrial control system test. Especially for the coverage of boundary conditions, abnormal scenes and concurrent paths, the manual dependence is reduced, and the reliability and consistency of the test are improved.

The invention discloses a multi-dimensional intelligent analysis and fault traceability system for ACU final test data, and relates to the technical field of data analysis. The multi-source heterogeneous data acquisition module is used for acquiring operation parameters, function test data and assembly data through a CAN bus, test equipment and a production system, and is provided with a high-precision timestamp; the data cleaning and preprocessing module is used for standardizing data, filtering and denoising, synchronizing time and complementing missing values; the multi-dimensional feature extraction module is used for extracting features from a time domain, a frequency domain, a time sequence and a space; the intelligent fault diagnosis module is used for classifying faults by using a random forest and a D-S evidence theory and evaluating severity; the fault traceability reasoning module is used for constructing a fault propagation graph and positioning root causes through a Bayesian network; and the visual display and early warning module is used for generating a three-dimensional fault tree and a thermodynamic diagram and triggering graded early warning. According to the invention, the fault detection rate and traceability efficiency are improved, the false alarm rate is reduced, early warning is realized, the ACU test process is optimized, and the product quality and safety are guaranteed.

The invention provides a simulation test system simulation degree evaluation method and device, equipment, a storage medium and a product. The method comprises the steps of determining a test scene corresponding to a to-be-tested vehicle according to a design operation range of the to-be-tested vehicle; according to the test scene, determining a corresponding evaluation item, and respectively testing in the simulation test system and the actual environment to obtain simulation test data corresponding to the simulation test system and actual test data of the to-be-tested vehicle in the actual environment; based on the simulation test data, obtaining a simulation evaluation index corresponding to each evaluation item; based on the actual test data, obtaining an actual evaluation index corresponding to each evaluation item; and obtaining a simulation degree evaluation result of the simulation test system based on the simulation evaluation index and the actual evaluation index. According to the method, the simulation data and the actual data are compared according to the scene characteristics and the evaluation indexes related to the driving scene, and targeted evaluation results can be accurately obtained in different driving scenes.

The invention discloses an aerospace equipment data-real fusion performance test method and device based on reinforcement learning, and the device comprises a digital twin model building module which is used for building an aerospace equipment high-fidelity mechanism model fusing multi-physical field features, a lightweight digital twinborn body adaptive to reinforcement learning training is generated through a model order reduction technology; the digital twinborn model correction module is used for carrying out consistency calibration of multi-dimensional features and key physical parameter correction on the reduced-order model based on object reference data; the performance test module based on reinforcement learning actively explores a coupling scene of an extreme working condition and multiple faults in an interaction environment by constructing an attacker agent, and autonomously explores a performance boundary and a limit envelope of equipment; and the number-real fusion verification module is used for carrying out a hardware-in-the-loop point compensation test by taking the performance limit obtained by the digital test as guidance so as to assist in decision making and optimization of a test scheme. According to the method, the coverage, the safety and the test efficiency of the limit performance test of the aerospace equipment are improved.

The invention discloses a method and device for automatically generating a Java method unit test case based on AI, and the method comprises the steps: building a structured annotation specification, carrying out the automatic traversal and intelligent analysis of a Java project, extracting a method signature, a parameter type and a parameter constraint condition, constructing a test data generation model, and carrying out the automatic generation of the test case. The model is transmitted to an AI service to analyze business logic and boundary conditions, test case codes are generated, test cases are optimized and integrated to a project test directory, finally, code areas with insufficient coverage are identified through coverage rate analysis, a test completion strategy is generated, and iterative perfection of the test cases is completed. The technical problems that in the prior art, project-level automatic test case generation cannot be achieved, the generated test case cannot fully cover business logic and boundary conditions, and context relation understanding is lacked are solved, full-automatic project-level test generation is achieved, and the test efficiency and quality are improved.

The invention provides a light-weight aerial ship image target detection method and device based on yov8n and a storage medium. According to the scheme, the method comprises the following steps that an image data set of an aerial ship is obtained and preprocessed; the method comprises the following steps of: constructing a network model of the improved yov8n, replacing part of an original CBS module with a GAC module, replacing part of an original C2f module with a C2fADWC module, modifying a detection head and a loss function, and selecting a rotating frame OBB; the training set is input into the improved yolov8n-obb model to be trained; a trained yolov8n-obb improved model is obtained; and inputting the test data set into the trained improved lightweight yolov8n-obb model to obtain a ship detection and identification result. According to the method, the ship in the aerial ship image can be more accurately identified, the wharf and the ship can be better distinguished, and the detection precision is remarkably improved while extremely little parameter quantity is increased.

The invention discloses a test data intelligent generation and privacy compliance desensitization method, which belongs to the technical field of electric digital data processing, and comprises the following steps: monitoring the structure change of a preset regulation update source and a target application system, and obtaining compliance dynamic information; analyzing the compliance dynamic information to generate a dynamic compliance rule base; based on the dynamic compliance rule base, generating an initial test data set through a preset intelligent data generator; desensitizing the initial test data set, and outputting a desensitized test data set; scanning the desensitized test data set in real time for compliance verification, and generating a compliance verification report; and dynamically optimizing the generation strategies of the dynamic compliance rule base and the intelligent data generator based on the compliance verification report. Dynamic rule base construction, intelligent separated data generation, closed-loop verification optimization and desensitization audit tracing technologies are adopted, laws and regulations and system changes can be automatically adapted, and the service validity, privacy security and law compliance of test data are synchronously guaranteed.

The invention discloses a method, a system and equipment for predicting the stability of integrated circuit test equipment and a medium, and belongs to the technical field of integrated circuit test. The method comprises the following steps: firstly, carrying out preprocessing and feature selection on historical data in an FT test stage, and adopting a Gaussian mixture model (GMM) to cluster and identify different operation condition clusters of a test machine; then, establishing a health state GMM reference for each working condition cluster, calculating a KL divergence value of a normal sample and the reference, and setting a dynamic anomaly detection threshold by 99.73% quantile of the KL divergence value; and finally, in real-time monitoring, calculating a KL divergence value of real-time data and a corresponding working condition cluster benchmark, and comparing the KL divergence value with a dynamic threshold value to realize accurate anomaly marking. The method effectively solves the problem of abnormal detection of the test data of the integrated circuit under complex and changeable working conditions, and improves the monitoring accuracy and working condition adaptability.

The invention belongs to the technical field of semiconductor defect detection, and relates to a deep learning-based deep energy level defect identification method, system and device, and a medium. According to the method, transient curve data of an external capacitor is modulated to obtain a time sequence feature vector containing time, temperature and bias voltage information, then a multi-defect-induced weighted capacitor feature vector is obtained through calculation and modulation of a neural network model, and the time sequence feature vector and the multi-defect-induced weighted capacitor feature vector are trained after being spliced. Until the neural network model converges, the trained neural network model predicts a defect parameter vector, defect weight distribution and transient capacitance response at the (i + 1) th moment at the ith moment; according to the method, DLTS test data is modeled and analyzed in combination with a deep learning algorithm and a semiconductor deep energy level defect kinetic model, precise recognition and quantification of a complex defect system are achieved, the problem that defect recognition precision is insufficient in an existing DLTS test method is solved, and meanwhile the method is low in requirement for test equipment and high in applicability.

The invention provides an intelligent fault diagnosis method and system for electromagnetic valve test data and a computer program product, and relates to the technical field of electromagnetic valve fault diagnos.The intelligent fault diagnosis method comprises the steps that firstly, multi-dimensional test data of an electromagnetic valve under multiple working condition combinations are collected, and multiple types of data including action response and the like are contained; a data behavior collaboration model is constructed based on the behavior collaboration relation between the data, and laws such as synchronous response are embodied; an abnormal collaboration chain deviating from a normal collaboration mode is mined from the model, and abnormal performance of each link is determined; reversely deducing a fault core cause through the abnormal collaborative chain, and determining information such as cause types; and finally, generating an executable instruction sequence based on the fault core inducement and the abnormal collaborative chain conduction characteristic, and controlling the maintenance system to execute fault handling operation. Therefore, faults can be comprehensively and accurately diagnosed, and the operation reliability and the system stability of the electromagnetic valve are improved.

The invention relates to the technical field of data storage, and discloses a solid state disk total write-in data volume testing method, which comprises the following steps of: initializing a testing environment and self-adaptive parameter configuration of a solid state disk; creating a plurality of parallel data acquisition threads, acquiring test data through the data acquisition threads, and storing the acquired test data into a database; establishing a preset health degree evaluation model, analyzing the collected test data in real time according to a preset health degree index system, dynamically adjusting a test load based on a dynamic acceleration stress model and according to the real-time state of the solid state disk, executing a corresponding test operation according to an adjustment result, and recording a test result; and analyzing the test result, and automatically generating a test report according to the analysis result. According to the method, the dynamic acceleration stress model is introduced, so that the test load is dynamically adjusted according to the real-time state of the SSD, and the test efficiency and accuracy are improved.

The invention relates to the technical field of power grid operation and maintenance, and provides a generator set operation performance processing method and device, a generator set operation performance processing system, a readable storage medium and a program product. The method comprises the following steps: fusing a static parameter sequence, a historical operation data sequence, a real-time monitoring data sequence and an active disturbance test data sequence of the generator set to obtain a fused input vector sequence; according to the fusion input vector sequence and the generator set performance evaluation model, obtaining an operation performance prediction index value sequence; according to a set length time window and a set sliding step length, sliding on the operation performance prediction index value sequence to obtain a target operation performance prediction index value sequence; and determining a credibility first score, a drift statistic and a test statistic of each target operation performance prediction index value sequence so as to judge whether to correct or retrain the generator set performance evaluation model. By adopting the method, high-precision prediction of the operation performance of the generator set can be realized.

The invention provides a performance testing method and device for a large language model and related equipment. The performance testing method comprises the steps that multiple testing platforms for testing, testing scene configuration parameters and a structured testing data set for testing are determined; on the basis of the test scene configuration parameters and the structured test data set, synchronous constraint parameters are generated and distributed to the multiple test platforms, and then a preset load generator is used for controlling the multiple test platforms to concurrently execute reasoning test tasks about the large language model; and when the test is executed, collecting multiple dimension performance indexes of each test platform in real time for data analysis, and generating a reasoning performance comparison evaluation report of the large language model in the multiple test platforms according to a data analysis result. According to the method, the cross-platform performance evaluation period is greatly shortened, and the waste of computing resources is reduced. Manual operation errors are avoided, comparability of multi-platform performance indexes is guaranteed, and more accurate and reliable multi-platform performance test results can be obtained.

The invention provides a fretting fatigue testing device and method in an atmosphere environment, and relates to the technical field of material performance testing in a complex environment, and the device comprises an upper clamp, a lower clamp and a sealing cavity; the upper clamp and the lower clamp are respectively fixed in an upper chuck and a lower chuck of the first load loading mechanism; the sealing cavity is used for providing a sealed service atmosphere environment for a fatigue specimen and comprises two flange end covers and a tubular connecting piece, the two flange end covers are respectively connected with the upper clamp and the lower clamp in a sealing manner; the two ends of the tubular connecting piece are in sealing connection with the upper end cover and the lower end cover respectively, and sealing connection of at least one end is of a detachable structure. The device can realize accurate simulation and steady state maintenance of service gas, is suitable for multiple scenes such as dry / wet air, inertia, oxygen / hydrogen enrichment and corrosive gas, and overcomes the limitation that traditional atmospheric environment test data cannot be used for real life prediction.

The invention provides a rock mineral analysis test management method and system, and the method comprises the steps: collecting rock mineral sample multi-modal data, and constructing a digital twinborn model; constructing a geological knowledge map, and injecting and fusing the two knowledge to generate a virtual test data set; performing self-supervision pre-training by using the data set to obtain a self-adaptive mineral phase recognition model; deploying the model to a block chain-Internet of Things double-chain traceability system to obtain an identification result and an optimization instruction; and executing a sample preparation process and removing unqualified samples. According to the method, diffusion steps are adjusted through adaptive noise scheduling and a space-time alignment algorithm, data are integrated, a high-reality-sense virtual test data set is generated, and a self-supervised pre-training and knowledge distillation technology is combined, so that the model has space-time feature extraction and dynamic optimization capabilities; and finally, full-process closed-loop control is realized through a double-chain traceability system and a modular robot, and the recognition precision of a complex scene is improved.

The embodiment of the invention discloses an evaluation model training method. According to the method, a standard operation program text is disassembled into function points, then each function point is realized in a code mode and in a mode of calling a large model, the two realization modes are tested by utilizing test data, and applicability indexes of the two realization modes to the function points are determined based on test results. And utilizing the applicability index data to train an evaluation model which can quantitatively judge whether the function point is suitable for being realized by using a code or a large model according to the description of the function point. The evaluation model training device and the data processing method and device provided by the embodiment of the specification also have the above beneficial effects.

The invention discloses a POL power supply automatic test method based on modular design, and belongs to the technical field of power supply test. The method comprises the following steps: firstly, under the control of an upper computer, automatically reading POL power supply basic parameters including key characteristic parameters such as input / output voltage, rated current and frequency from a preset Excel table, and establishing a judgment reference for testing; then, instrument initialization is completed, communication connection with test equipment such as an oscilloscope, an electronic load and a direct-current power supply is established, and configuration of initialization function parameters is completed; then, executing a measurement task of each core parameter of the POL power supply to be tested, and in the test process, executing a parallel measurement management mechanism, namely configuring each instrument parameter and channel state in real time according to a specific test item, and triggering a retry or interrupt logic under the conditions of communication abnormality, data timeout and the like until the test recovers a running state; and finally, after the test is completed, performing Excel storage and interaction on the standardized test data and the waveform screenshot, and finally safely exiting. According to the POL power supply automatic test method based on the modular design, the whole test process is divided into a plurality of serial or parallel function processing units such as parameter reading, instrument initialization, test execution and test management, data storage and interaction and the like, and high efficiency and usability of a test system are achieved.

The invention provides an aero-engine test whole-process business management method and system, and belongs to the technical field of whole-process business management, and the method comprises the steps: extracting information in a test item information document and a tester information document, and determining structured data; on the basis of the structured data, forming an initial test plan, performing test personnel allocation optimization on the initial test plan, and determining a target test plan; generating a target test card on the basis of the target test plan, and continuously updating the target test card by adopting an editable block chain according to the data in the test process until the test is finished; and after the test is finished, the operation authority of the target test card is removed, and non-tampered and decentralized aero-engine test whole-process business data is formed. The information utilization efficiency of each link is effectively improved, the distribution rationality of testers is improved, and the safe storage of test data is ensured.

The embodiment of the invention discloses a test case generation method and device based on a service scene and related equipment. The method comprises the following steps: acquiring historical test data, and identifying and extracting entities and corresponding entity relationships in the historical test data through a pre-trained identification model; according to the entities and the corresponding entity relationships, storage is carried out in a triple form, and a test knowledge graph is constructed; analyzing based on the test document of the current test item, and determining the test requirement and the test item of the test document; based on the test requirements and the test items, matching corresponding standard triads and reference test cases from the knowledge graph and a preset test case library to obtain a target triad set and a target reference case; and generating a target cue word based on the test requirement, the test item, the target triple set and the target reference case, and generating a target test case through a pre-trained large language model. According to the method, the scene coverage of the test case is higher, and the logic continuity of the test case is improved.

The invention relates to a method, a system and equipment for predicting creep stress at a joint of a concrete structure and a medium. The method comprises the following steps: testing a concrete structure material and a concrete structure joint filling material, and constructing a joint material performance database based on test data; constructing a multi-scale mechanical model comprising a macrostructure model and a microcosmic interface model based on the seam material performance database; constructing an interface constitutive relationship based on the seam material performance database and the multi-scale mechanical model; and predicting the creep stress at the joint of the concrete structure based on the material performance database at the joint, the multi-scale mechanical model and the interface constitutive relationship. The method can fully consider the sudden change of the performance of the material at the joint and the interface effect, provides reliable technical support for the safety evaluation and maintenance of the concrete structure, and is particularly suitable for engineering application of large-span and high-rise buildings and the like with higher requirements on the long-term performance of the structure.

The invention discloses a multifunctional hydraulic system experiment table, which relates to the technical field of hydraulic system experiment tables and comprises a time base line synchronous control module, a phase lock correction stability control module, a false peak identification and correction module, a multi-dimensional fault-tolerant judgment control module and a reverse energy slow-release anti-reverse module. And the time base line synchronous control module is used for establishing a cross-sensor unified time base line control structure, generating a synchronous reference surface by using a high-stability crystal oscillator signal, and executing nanosecond phase calibration on sampling signals of the pressure sensor and the flow sensor by taking the synchronous reference surface as a unified time reference. According to the invention, high-precision synchronous acquisition of pressure and flow signals is realized through unified time base line and phase locking correction, and high-pressure transient test data is ensured to be real and stable; and meanwhile, control logic is reconstructed, and an energy slow release strategy is introduced, so that the pressure relief process is stable and controllable, pump reverse rotation and structural impact are avoided, and the safety, reliability and self-healing stability of the experiment table are remarkably improved.

The invention discloses a concrete nonlinear creep explicit modeling method and system, and relates to the technical field of concrete structures and materials.The concrete nonlinear creep explicit modeling method comprises the steps that concrete creep test data under all conditions are collected, and a test data set is constructed; preprocessing the concrete creep test data of the test data set; a DBO algorithm is adopted to optimize hyper-parameters of the LGBM model, and the LGBM model is constructed and trained through the test data set and the optimal hyper-parameters; the input features of the trained LGBM model are screened through an SHAP additive interpretation method, and key feature parameters are obtained; and establishing a concrete nonlinear creep explicit calculation formula by using a PySR symbol regression tool and taking the key characteristic parameters as input and the creep strain as output. The concrete nonlinear creep explicit modeling method provided by the invention has both calculation precision and engineering practical value.

The invention relates to the technical field of model testing, in particular to a model testing method and device and electronic equipment, and is used for solving the problems of long test data combination time consumption and low coverage rate of a model testing mode in the related technology. The multiple features are classified, a core feature set, a common feature set and an edge feature set are obtained, and the model can be a deep learning model; generating a test set of the model according to a differentiated data generation strategy of feature importance on the basis of the rule set of the service type adapted to the model, the core feature set, the common feature set and the edge feature set, and performing model performance test on the model by using the test set to obtain a test report; therefore, the efficiency of model algorithm testing is improved.

The invention provides an ultrasonic radar debugging data interaction method and system, and the method comprises the steps: carrying out the analysis of an obtained structured configuration file, and obtaining a parameter configuration set which comprises radar detection parameters and communication protocol parameters; generating an interaction control on the interaction panel based on the parameter configuration set, and establishing a bidirectional data binding channel between the interaction control and the parameter configuration set; in response to a parameter change operation of a user, extracting a change field of the parameter configuration set, and generating a difference configuration instruction based on the change field; driving an ultrasonic radar test based on the operation parameters of the dynamic parameter storage area, and collecting original signal data and processing result data to obtain a test data set; and performing association analysis on the difference configuration instruction and the test data set to generate test result data. By adopting the method, the efficiency and accuracy of functional verification of the radar sensor can be improved.