665 results about "Test analysis" patented technology

The invention relates to the technical field of automatic report generation, and provides a test analysis and report generation method and system based on large model retrieval enhancement, and the method comprises the steps: firstly carrying out the semantic task decomposition of user query, and meanwhile, achieving the multi-dimensional information extraction through the butt joint of a knowledge vector library and a structured image-text knowledge system established by a private knowledge base module. And three strategies of RAG, Self-RAG and Graph-RAG are combined to enhance the generation capability of the large model so as to accurately obtain background knowledge. And constructing a standard SQL statement according to a structured query requirement, and performing data filling according to template prompt by an output result fusion module in combination with a query background and an SQL execution result to form a test report. By optimizing a retrieval enhancement generation method, professional knowledge supplement related to query is realized, and the reasoning ability of a large model in a professional scene is effectively enhanced, so that the accuracy and the intelligent level of data analysis are improved.

The invention relates to the technical field of network service testing, in particular to a testing method and system based on optical fiber characteristics. The method comprises the following steps: acquiring multi-dimensional optical fiber characteristic data of an optical fiber link through optical fiber characteristic acquisition equipment; performing characteristic data hierarchical integration on the multi-dimensional optical fiber characteristic data to obtain an optical link characteristic vector group; performing temperature-error code wavelet transform association according to the optical link feature vector group to obtain a temperature-error code wavelet feature set; constructing an association prediction matrix according to the temperature-error code wavelet feature set; performing a service delay exception test according to the association prediction matrix to obtain a delay tail exception event set; and performing micro-bend three-dimensional feature recognition according to the delayed tail abnormal event set to obtain micro-bend mode features. According to the invention, the operation and maintenance efficiency, reliability and toughness of the optical fiber network of the data center are improved through full-link test analysis from optical fiber physical characteristics to storage service delay.

The invention discloses a ship mobile energy storage system CAN communication protocol automatic test system and method, and the system comprises a protocol modeling and test configuration module, a simulation node behavior simulation module, a CAN bus interaction and power closed loop module, and a test analysis and result evaluation module. A CAN communication protocol specification of the ship mobile energy storage system and parameters of tested equipment can be obtained, a protocol analysis model and a test configuration file are constructed for format verification and logic consistency verification, meanwhile, a corresponding test case sequence is automatically generated, and the message sending time is controlled to generate a time sequence simulation message flow; and sending the time sequence simulation message flow to the tested equipment, simulating energy interaction behavior acquisition to generate a power feedback data set, testing to obtain a signal consistency deviation and a power execution error to construct a protocol conformity evaluation matrix, and generating a visual test report. According to the invention, the high coverage rate and automatic closed-loop test of the CAN communication protocol can be realized.

The invention relates to the technical field of data testing, in particular to an intelligent test analysis system and method for an integrated circuit. The method comprises the following steps: acquiring circuit monitoring data; performing data preprocessing on the circuit monitoring data to generate circuit monitoring preprocessing data; performing multi-dimensional feature extraction on the circuit monitoring preprocessing data to generate a circuit monitoring multi-dimensional feature data set; performing feature set fusion on the circuit monitoring multi-dimensional feature data set to generate a circuit monitoring feature fusion data set; therefore, the problems of insufficient precision, response delay and high maintenance cost in traditional circuit monitoring are solved through an accurate data processing and intelligent analysis method, and the real-time performance, the accuracy and the predictability of intelligent testing of the integrated circuit are improved.

The invention relates to the field of chip testing, in particular to a chip signal integrity test analysis method, device and equipment and a storage medium. The method comprises the following steps: extracting a chip performance index based on a chip design requirement document; performing saturation frequency excitation setting based on the chip performance index to generate an excitation test signal; performing a multi-port signal excitation test and parallel signal sampling based on the excitation test signal to generate a high signal-to-noise ratio response data set; obtaining a chip layout design file; performing chip communication link analysis on the chip layout design file, and constructing communication line layout data; performing signal mapping and waveform quality evaluation on the communication line layout data based on the high signal-to-noise ratio response data set to obtain a node signal distortion rate; and performing signal integrity quantization and theoretical signal integrity score comparison based on the node signal distortion rate, and marking a fault node. According to the invention, the chip testing efficiency is improved, and the signal abnormal part of the chip is accurately and efficiently detected.

The invention discloses a code generation and restoration method and device based on multi-round feedback of a large language model, and the method comprises the following steps: 1, converting problem description into a structured tetrad (target, input, output and constraint) through a two-stage task demand deconstruction mechanism, and generating a pseudo-code frame according to the structured tetrad; 2, generating a plurality of candidate code schemes in parallel based on different implementation strategies; 3, selecting an optimal code implementation scheme through multi-dimensional evaluation; 4, generating an extended test case through boundary condition analysis and public test reasoning, and expanding a verification range; 5, classifying error types in combination with static analysis and dynamic execution information, and applying a special repair strategy; 6, providing detailed repair reasons and ideas by an interpretable repair mechanism; and 7, establishing a multi-round feedback repair verification iteration mechanism composed of testing, analysis, repair and verification. According to the method, the performance of an existing large language model in processing a complex programming task is effectively improved.

The invention discloses an intelligent flow sensor test analysis method and system, and relates to the technical field of sensor testing, and the method comprises the steps: applying a stepped flow excitation signal to a tested sensor through a standard flow generation device, and synchronously collecting the dynamic response data of the sensor; extracting a feature parameter set from the response data; performing similarity matching calculation to obtain a matching degree evaluation value of each error type; quantifying the induction relation strength among different error types; generating a composite weight coefficient in combination with the matching degree evaluation value and the evoked relationship strength; and making a multi-stage compensation strategy based on the composite weight coefficient. The method has the advantages that by combining error correlation network modeling driven by historical data, induction mechanisms and interaction relations of different error types can be analyzed, the pertinence of a compensation strategy is greatly improved, a multi-stage error suppression mechanism is formed, stable control over comprehensive errors is achieved under complex working conditions, and meanwhile the system maintenance requirement is lowered.

The invention provides an integrated circuit performance testing method based on multi-dimensional parameter fusion, and belongs to the technical field of integrated circuit performance testing. S2, collecting multi-dimensional parameters; s3, data preprocessing; s4, carrying out multi-dimensional parameter fusion; and S5, performance test analysis. Through multi-parameter coverage, multi-algorithm collaboration and full-process closed loop, the problems of isolation, fragmentation and simplification of traditional testing are solved, and systematic and refined technical support is provided for performance testing of the integrated circuit.

The invention discloses a new energy broadband test analysis management and control method and system, and relates to the technical field of new energy detection and intelligent management and control, and the method comprises the following steps: obtaining a multi-window splicing frequency spectrum of a load switching edge, superposing phase lock consistency verification in a splicing frequency spectrum result, generating a frequency spectrum leakage suspicion chart, and carrying out the boundary constraint for energy track extraction; under the constraint of a spectrum leakage suspicion chart, an energy center track, a group delay curve and a phase derivative sequence are extracted, and a cross-frequency migration fingerprint is constructed and is used as a basic evidence for misjudgment traceability. By means of spectrum suspicion identification, feature fusion modeling, misjudgment traceability, label correction, energy reverse intervention and the like, accurate identification and closed-loop regulation and control of false operation caused by spectrum leakage are achieved, high identification accuracy, low misjudgment rate, high anti-interference performance and quick response capacity are achieved, and the operation stability and safety of a new energy field station are remarkably improved.

The invention is applicable to the technical field of PCB testing, and particularly relates to a PCB electrical performance testing method, which comprises the following steps of: acquiring test object information; wherein the object information comprises a PCB image and at least one test area, the PCB image corresponds to the PCB and reflects the circuit structure of the PCB, the test area is marked on the PCB image, and the test area comprises a main test position and an auxiliary test position; when an electrical performance test instruction is received, contact judgment information is obtained based on the test object information; when it is detected that the auxiliary test probe is in stable contact with the auxiliary test position, the control device controls the main test probe to continuously send test signals and collect test results in real time, and test analysis information is obtained based on the test results. According to the PCB electrical performance test method provided by the embodiment of the invention, the problem of high test error caused by asynchronous actions of the main test probe and the auxiliary test probe can be improved.

The invention relates to the technical field of chip test analysis, in particular to an analysis optimization method and system based on chip test data, and the method comprises the following steps: carrying out the voltage collection and current collection based on the chip test data and probe contact information, carrying out the node relation operation through combining with a graph neural network, and extracting the adjacency relation; and abnormal waveforms and noise waveforms are screened, wafer atlas positions are associated, and channel waveform mapping is generated. According to the method, voltage and current data and probe contact information are integrated, the graph neural network is adopted to deeply explore the relationship between nodes, the adjacency relationship is effectively extracted, the abnormal waveform is screened and associated to the wafer graph position, and the analysis capability of waveform data is improved. According to the strategy, by dynamically comparing the waveform difference values and counting the waveform amplitude and optimizing the test steps through the deep reinforcement learning technology, the efficiency of the test process is effectively improved, and redundant operation is reduced.

The invention relates to the technical field of hydrogen energy safety, and provides a hydrogen fuel cell automobile hydrogen leakage diffusion concentration monitoring test device and method.The device comprises a hydrogen storage system, a flow control system, a data acquisition and storage control system and a leakage opening simulation device, the hydrogen storage system comprises a hydrogen cylinder, and the flow control system comprises a plurality of airflow channels arranged in parallel; one end of each gas flow passage is connected with the hydrogen storage system, the other end of each gas flow passage is connected with a leakage port of one leakage port simulation device, each gas flow passage is provided with a plurality of gas flow branches in parallel, and different gas flow branches regulate and control the leakage gas flow by controlling the gas flow or the pressure and the size of the leakage port in a combined manner; according to the automobile hydrogen leakage diffusion test system, the hydrogen leakage diffusion platform of the hydrogen fuel cell automobile and the data acquisition and storage control system are built, so that the diffusion characteristics of the automobile after hydrogen leakage can be tested and analyzed, and the result can assist in making a hydrogen leakage emergency scheme and preventing the occurrence of hydrogen leakage fire explosion.

The invention relates to the technical field of network security, and discloses a multi-agent cooperative RESTful API vulnerability test method. Constructing a parameter structured representation by analyzing an API interface specification, generating a test value conforming to service semantics by using a large language model, and establishing a semantic request baseline representing normal service logic; constructing a multi-agent collaborative architecture including vulnerability reasoning, case generation, test execution, test analysis and test management; all the agents are coordinated to carry out iterative vulnerability testing through a signal mechanism and a state transition table, and API service analysis, test strategy customization, use case directional variation and execution, result study and judgment and strategy optimization are achieved; and finally generating a structured vulnerability report. According to the method, the problems of semantic blind areas, shallow analysis layers, insufficient adaptive capacity and the like are solved, security risks such as business logic vulnerabilities and sensitive data leakage in the RESTful API can be deeply detected, and the accuracy and coverage rate of vulnerability detection are remarkably improved.

The invention relates to the technical field of data analysis, in particular to a method for constructing an acute pancreatitis severity prediction model based on a liquid neural network, which comprises the following steps: firstly, extracting and preprocessing patient data, and then dividing AP patients into a severe AP group and a mild AP group; screening two groups of characteristic indexes with differences through nonparametric test analysis to carry out machine learning modeling in the next step; in order to prevent overfitting, five-fold cross check is adopted to train the model, and optimal feature combinations under respective machine learning models (a liquid neural network, a decision tree, a random forest, Logistic regression and an XGBoost algorithm) are obtained by drawing AUC values under an ROC curve for comparison. The method comprises the following steps: firstly, training a liquid neural network model, then verifying verification data by using respective optimal prediction models obtained by training, and finally, analyzing influence weights of different features in the liquid neural network model through SHAP visualization and carrying out visualization so as to provide visual prediction reference for users such as doctors and the like.

The invention provides a method for recovering free hydrocarbon contents of different lithologic shale. The method comprises the following steps: carrying out closed freezing heat release, freezing pyrolysis and freezing multi-temperature-order pyrolysis experiment test analysis on a closed coring well core; performing freezing pyrolysis experiment test analysis on the core of the non-closed coring well; carrying out pyrolysis experiment test analysis on the multi-courtyard core; determining the lithology type of the shale; and recovering the free hydrocarbon content of the shale according to an experimental test analysis result and the lithology type. According to the method, light hydrocarbon loss can be reduced to the greatest extent, the content of free hydrocarbons in different lithology non-closed coring well samples and placed samples is recovered, and the accuracy of shale oiliness evaluation is improved.

The invention discloses a system and a method for testing and analyzing the single event effect of a high-capacity nonvolatile memory. The system comprises an upper computer module, an FPGA (Field Programmable Gate Array) control module and a programmable power supply module. The system is used for testing and analyzing various single event effects of a high-capacity nonvolatile memory chip in a dynamic working mode. And according to the error data read from the memory in the dynamic working mode, presetting error bitmap rules of various single-particle function interruptions in an FPGA test program, carrying out real-time analysis and identification on the various single-particle function interruptions, and discriminating single-particle upset of the memory unit. According to the system, testing and real-time analysis of various complex single event effects of the high-capacity nonvolatile memory can be achieved, then radiation hardening and evaluation can be conducted on the sensitive position of a memory chip, and high pertinence is achieved.

The invention relates to the technical field of test analysis, and discloses a probe device for plasma parameter measurement, which comprises a probe assembly, a cooling assembly, a heat dissipation seat and a fixed seat, a cooling cavity in a heat sink column connected with a probe and a heat dissipation channel in a heat dissipation base are connected with a water inlet pipe and a water outlet pipe respectively, so that heat dissipation and cooling can be achieved through the cooling cavity and the heat dissipation channel when the probe is subjected to heat load from plasma, accumulated heat of the probe is taken away, the heat dissipation efficiency is improved, and the service life of the probe is prolonged. Therefore, the probe can bear a thermal load for a longer time, so that the normal operation of the measurement work is ensured. Besides, by arranging the at least two probes, parameter measurement can be carried out on at least two positions at the same time, and the probes can also be integrated for measurement, so that part of system errors can be eliminated or the central position of the plasma can be found through the collected at least two groups of parameters, and a more accurate measurement result can be obtained.

The invention provides a mining motor no-load and load characteristic test analysis method and system, and relates to the field of intelligent image analysis, a Hall array sensor is arranged on a motor stator tooth portion to collect air-gap magnetic field distribution data, and a high-frequency signal injection method and a wavelet singular value decomposition technology are combined, so that the load characteristic of a mining motor is obtained. The air gap magnetic field distortion degree and the air gap dynamic eccentricity index are accurately calculated, and meanwhile a load dynamic compensation coefficient is generated to correct the motor output torque. Therefore, the dynamic characteristics of the motor in the no-load and load states can be comprehensively reflected, the accuracy of the load capacity evaluation of the mining motor is improved, and potential problems in the operation of the motor can be found in time, so that the performance of the motor is optimized, the service life of the motor is prolonged, and the safety and reliability of mining production are improved.

The invention provides a marine environment steel structure coating anticorrosion performance test analysis method, which comprises: presetting marine environment data and coating information data under different conditions, adopting an accelerated corrosion test to test steel structure coating samples under different conditions, and collecting coating performance time sequence change data under different test conditions; processing the coating performance time sequence change data under different test conditions, determining a coating performance effective time point, extracting and calibrating time sequence characteristic information from the coating performance to the effective time point, and establishing a sample training data set; constructing an anti-corrosion performance analysis model, inputting the sample training data set into the model for training, and outputting the residual life of the coating; the marine corrosion process is accelerated and simulated through multi-dimensional environmental parameter presetting and orthogonal test design, and the residual life of the coating is accurately predicted by combining a collaborative test of electrochemistry, mechanics and microtopography and constructing an anti-corrosion performance analysis model, so that subsequent coating maintenance and management are facilitated.

The invention relates to the technical field of server aging testing, in particular to a server aging testing method and device, electronic equipment and a storage medium. Then convolution processing is carried out on each aging data queue in a sliding mode through a difference operator, the obtained multiple first convolution queues are segmented one by one, queue segments obtained through segmentation are normalized, and multiple queue segments are obtained; performing dot product calculation on each queue segment, each same-time-segment queue segment and each same-part queue segment to obtain a plurality of dot product results, and adding representative values determined according to the plurality of dot product results into a test analysis matrix; and finally, according to the plurality of row vectors of the test analysis matrix, determining the abnormal probability of the plurality of tested parts of the server. According to the method, the state of the detected component is determined based on the cross point product and probability estimation, so that the problem component can be found early.

The invention provides a new energy inverter aging test method and system, and relates to the technical field of inverter testing, and the method comprises the steps: building an inverter aging test platform; performing test parameter analysis based on the attribute information of the inverter and the test application demand, and generating an aging test parameter table; performing test optimization and monitoring perception on the target inverter according to the aging test parameter table to obtain an inverter aging test multi-dimensional data stream; key feature extraction and attenuation degree evaluation are carried out on the aging test multi-dimensional data flow, and an inverter operation performance attenuation parameter set is obtained; and calling an aging performance analyzer to carry out aging performance evaluation on the operation performance attenuation parameter set, and determining an evaluation result. According to the aging test method and device, the technical problem that the test result is lack of pertinence and accuracy due to the fact that the test parameters are fixed and customized test analysis under different performance characteristics and application conditions is lacked in the prior art is solved, and the pertinence, adaptability and accuracy of the aging test are improved.

The invention discloses a biaxial test device, comprising: a bracket, a container, a sample setting assembly and a press, etc. Before use, the sample is installed between two limit plates. On the basis of applying confining pressure, the press is started, and the pressure rod moves vertically downward and applies a vertical downward load to the sample. When the sample is subjected to the vertical downward load, axial compression and radial deformation will occur. At this time, the radial deformation of the sample on both sides of the limit plates is limited, and expansion occurs on both sides without the limit plates. Subsequently, under the joint action of the confining pressure and the limit plates, the sample undergoes plane strain. Since the container and the limit plates are both in a transparent state, the side section state of the sample when plane strain occurs can be well observed. At the same time, the multiphase mixed fluid in the mixing container penetrates into the rock sample inside the sample and flows out through an output pipeline and is tested and analyzed by external equipment, so that the potential influence of multiphase seepage on the mechanical behavior of rock mass can be studied under plane strain conditions.

The invention relates to a driver fatigue state data determination method and device and computer equipment. The method comprises the following steps: acquiring environment state data corresponding to a running automobile and driver action data corresponding to a driver; according to the environment state data, risk events of the running automobile in a future preset time domain are predicted and analyzed, and task complexity analysis data are obtained; performing mutual information analysis on a corresponding relation between the environment state data and the driver action data to obtain driver-man mutual information analysis data; according to the task complexity analysis data and the driver mutual information analysis data, calculating a fatigue index of the driver at the current moment to obtain an information efficiency attenuation fatigue index; and performing sequential statistical test analysis on the information efficiency attenuation fatigue index to obtain fatigue state data corresponding to the driver. By adopting the method, the generalization ability and the stability of determining the fatigue state of the driver can be effectively improved.

The invention relates to the field of test and analysis of impurity element content of primary magnesium ingots, in particular to a method for testing impurity element content of primary magnesium ingots by ICP (inductively coupled plasma) spectrometry, which comprises the following steps: firstly adding H2O2 into a magnesium ingot sample to infiltrate the sample, then adding HCl at an interval of less than 1 minute to carry out first reaction, carrying out first heating after the reaction is finished, and then carrying out second heating, cooling to 20-30 DEG C, transferring to a volumetric flask, and fixing the volume by using third-level water to obtain a sample solution; establishing a working curve by using an ICP spectrometer and a standard solution; testing the sample solution; the ratio of the mass of the magnesium ingot sample to the volume of H2O2 to the volume of HCl is 1: (3-10): (30-54), the method fills the standard vacancy of ICP spectrometry analysis of the silicon element in the primary magnesium ingot, a reliable basis is provided for chemical component analysis of the primary magnesium ingot, the test result is surface, and the method for testing the mass fraction of the impurity elements in the primary magnesium ingot provided by the invention has the advantages of high analysis precision, simple operation, high accuracy and high reliability. The analysis period is short, and related chemical varieties are few.

The invention discloses a method and a device for determining a diffusion coefficient in a charged micropollutant nanofiltration membrane in the technical field of membranes. The method comprises the following steps: arranging a four-electrode electrochemical system in an H-type electrolytic tank, adding a micropollutant solution into the H-type electrolytic tank, and analyzing the micropollutant solution by adopting linear scanning voltammetry to obtain first data; according to the method, a micro-pollutant solution system, a micro-pollutant solution and nanofiltration membrane system and micro-pollutant solutions and nanofiltration membrane systems with different concentrations are tested and analyzed based on a four-electrode electrochemical system in an H-shaped electrolytic tank, and corresponding solution intrinsic conductivity data, system conductivity data and open-circuit voltage data are obtained; the method has the advantages of simple test steps, fast data acquisition, short overall test period, fast replacement and adjustment of the micropollutant solution and the nanofiltration membrane, wide application range, simple and effective linear fitting data processing process, and good repeatability.

The invention discloses a gallium oxide semiconductor device test data analysis system, which relates to the field of semiconductors, solves the problem that the existing semiconductor device test data analysis system is poor in analysis effect, and comprises an electrical index module, a data analysis module, a data analysis module and a data analysis module, the photoelectric index module is used for carrying out various photoelectric index tests on the sample semiconductor device in the photoelectric index test period and obtaining a periodic photoelectric index sorting queue according to the test result; and the test analysis module is used for carrying out test data analysis on the sample semiconductor device according to the periodic electrical index sorting queue and the periodic photoelectric index sorting queue, and feeding back an analysis result. According to the invention, the accuracy and comprehensiveness of the test data analysis result can be improved.

The invention belongs to the field of antenna quality detection, relates to a data analysis technology, is used for solving the problem that the electrical parameter performance of an antenna cannot be tested and analyzed in the prior art, and particularly relates to a dual-notch ultra-wideband antenna quality detection method and system. Comprising an electrical parameter test module, an environment adaptation test module and a comprehensive evaluation module which are connected in sequence, and the environment adaptation test module is further in communication connection with an adaptation optimization module; the electrical parameter test module is used for testing and analyzing electrical parameters of the antenna; the antenna is marked as a test object, and the test object is marked as an electrical parameter normal object or an electrical parameter abnormal object through a voltage standing wave ratio; according to the invention, the environmental adaptability of the antenna can be tested and analyzed, the electrical parameter performance of the single antenna is verified again according to the generated test environment, and the environmental adaptability of the batch of antennas is evaluated according to the proportion of the number of adaptive qualified objects in the verification result. And triggering adaptive optimization analysis when the environmental adaptability is abnormal.

The invention relates to the technical field of operation management, in particular to an airport intelligent global situation awareness and intelligence analysis management method and system. The method comprises the following steps: capturing corresponding global operation data in real time to generate an airport global original data set; performing airport operation map construction and pressure elasticity test analysis on the airport global original data set to generate an airport operation global pressure assessment result; performing airport operation pressure planning and global situation awareness comparative analysis on the basis of an airport operation global pressure assessment result to generate an airport actual operation global situation awareness model; carrying out risk early warning information analysis and airport omnibearing comparison based on the airport actual operation global situation awareness model, and generating an airport global operation process redisk report containing deviation analysis; and performing operation optimization and improvement analysis based on the airport global operation process recovery report to obtain an airport operation global optimization and improvement management scheme. The airport operation efficiency can be remarkably improved.

The invention provides an on-line power device aging test circuit facing vehicle working conditions, a comprehensive test analysis method and a comprehensive test analysis system. The aging test circuit comprises a tested three-phase six-bridge arm circuit composed of a tested power device, a temperature-sensitive electrical parameter measurement circuit, a bridge arm switching circuit and a standby bridge arm circuit. The temperature-sensitive electrical parameter measuring circuit is used for measuring the temperature-sensitive electrical parameter time-varying characteristic of the accessed tested power device; the bridge arm switching circuit is used for connecting the tested power device to the tested three-phase six-bridge arm circuit or the temperature-sensitive electrical parameter measuring circuit; the standby bridge arm circuit is used for replacing any bridge arm circuit where the tested power device connected to the temperature-sensitive electrical parameter measuring circuit is located so as to form a standby three-phase six-bridge arm circuit. The on-line power device aging test circuit provided by the invention can perform on-line detection on the aging condition of the vehicle inverter power device in a power cycle state.

The invention discloses a chip test retrieval method and system based on a large model, and the method comprises the following steps: collecting chip test documents, and constructing a chip test vector knowledge base; receiving user query, and executing search in the knowledge base by using the word vector model structure; richer context information is generated based on a retrieval result and user query, and a test case or a test analysis report is generated through a fine-tuned large model. According to the method, the retrieval capability and the generation capability are effectively integrated, efficient retrieval of test instructions and knowledge and automatic generation of test cases / reports are achieved, and the chip test efficiency and quality are remarkably improved. Meanwhile, the retrieval generation system adopts a Django framework development platform, performs real-time streaming output and real-time feedback of retrieval results by means of the MyWebsocket technology, and writes question and answer data into MySQL. The system comprises a knowledge base construction module, a retrieval module, a generation module, a streaming interaction module and a data storage module. Compared with the prior art, the method has the advantages that the test instruction retrieval accuracy is improved by more than 18%, the test case generation time is shortened by 60%, and the test knowledge is shared and reused.