A linked flexible test system

Abstract

The present application relates to a kind of linkage flexible test system.Therein, the system includes: data acquisition module, data processing module, multiple linkage resource scheduling module, flexible test scheme enabling module, intelligent control test report module;Realize the test system is used to build adaptive test environment, and the data of equipment under test is collected and generates multidimensional original matrix data;After obtaining the time series unified data of pre-processing, it is structured storage;Priority adaptation equipment list is used to determine the required equipment, and communication link is constructed by protocol adaptation dynamic decision tree, triggers fault reasoning model monitoring equipment initialization configuration, and realizes dynamic resource scheduling using multiple equipment scheduling algorithm;Flexible test scheme enabling module presets test step, rule and parameter, matches protocol analysis component, and generates automation script;Intelligent test report module captures data in real time by script, and according to preset rule, discrete data and demand index are mapped and matched, and multidimensional test result is generated.

Description

Technical Field

[0001] This invention belongs to the technical field of industrial internet, specifically relating to a linkage flexible testing system. Background Technology

[0002] The current flexible testing system still has the following areas for improvement:

[0003] The primary pain point for existing testing companies lies in the imbalance between testing resource utilization and response efficiency. Most testing companies have their equipment scattered across different laboratories or project teams, lacking a unified resource integration and sharing mechanism. This results in a persistent contradiction between partially idle and partially scarce equipment. Some specialized equipment serves only specific projects, becoming idle and wasted after project completion. However, when new projects are launched, equipment needs to be temporarily allocated or even repeatedly purchased, significantly reducing equipment sharing rates and overall testing capabilities. Furthermore, facing diverse customer testing needs, traditional fixed test benches cannot be quickly configured, requiring manual rebuilding of the testing environment and debugging of equipment parameters. This leads to long delivery cycles for testing solutions, hindering rapid response to market demands for testing efficiency and further weakening the company's competitiveness in the testing industry.

[0004] Insufficient test data quality and synchronization have become key pain points restricting the accuracy of testing companies. On the one hand, in existing testing processes, the lack of a unified, highly dynamic synchronization mechanism for multi-device data acquisition often leads to time differences in data collected from different devices (such as power supplies, loads, and environmental test chambers). Especially in highly dynamic testing scenarios, data asynchrony directly affects the reliability of test results and fails to provide accurate basis for performance evaluation of the tested products. On the other hand, some testing equipment is outdated or the data acquisition link is unstable, which easily leads to problems such as missing data and outliers. Manually completing data is inefficient and prone to introducing errors, further reducing the completeness and accuracy of test data, making it difficult to meet customers' core demands for high precision and high reliability of test results.

[0005] The lagging adaptability and intelligence level of testing scenarios limit the business expansion of testing companies. Traditional testing models rely on fixed testing procedures and manual operation. When faced with customers' personalized and customized testing needs (such as multi-module collaborative testing of new energy vehicle all-in-one controllers and special working condition testing of industrial equipment), they cannot flexibly adjust test variables, alarm thresholds, and process logic, resulting in poor universality of test solutions and requiring a large investment of manpower for customized development. At the same time, data recording, analysis, and report generation during the testing process largely depend on manual work, which is not only time-consuming and labor-intensive but also prone to deviations in analysis results due to human error. Furthermore, the lack of intelligent early warning and remote monitoring capabilities means that when equipment malfunctions or data exceedances occur during testing, manual on-site investigation and handling are required, increasing testing safety risks and management costs, making it difficult to adapt to the development trend of the testing industry towards automation and intelligence. Summary of the Invention

[0006] To address the aforementioned problems in the existing technology, this invention provides a linkage-type flexible testing system.

[0007] The objective of this invention can be achieved through the following technical solutions: a data acquisition module, a data processing module, a multi-linkage resource scheduling module, a flexible testing solution enabling module, and an intelligent control and testing report module;

[0008] The data acquisition module obtains the corresponding test environment based on the linked flexible test system, collects data from the device under test, and acquires multi-dimensional raw matrix data;

[0009] The data processing module verifies the multi-dimensional original matrix data through data integrity monitoring, triggers the value compensation function when the data is abnormal, and obtains the corrected data matrix; it performs data preprocessing on the corrected data matrix to obtain time-series unified data, and stores the data in a structured manner.

[0010] The multi-linkage resource scheduling module obtains a priority-ranked list of compatible devices and the test devices required for the test scenario based on the structured storage of the data. Based on the test devices, it constructs a communication connection link between the devices and the system through a protocol adaptation dynamic decision tree algorithm, triggers a preset fault reasoning model, monitors the initialization configuration of the test devices in real time, and performs multi-linkage resource scheduling through a device resource scheduling algorithm.

[0011] The flexible testing solution enabling module utilizes the multi-linkage resource scheduling to preset the execution order, condition judgment rules, and loop parameters of the test steps; at the same time, it connects to the built-in protocol parsing component required for testing and matches it with the communication environment to generate an automated script for the flexible testing solution.

[0012] The intelligent control and testing report module captures data in real time during the testing process based on the automated script; through preset data association rules, it maps and matches the captured discrete test data with test requirement indicators to generate multi-dimensional test results.

[0013] As a preferred technical solution of the present invention, obtaining a test environment of the corresponding scale includes, based on the number of devices under test, the type of test scenario, the density of test data acquisition, and the test requirements, activating the basic acquisition device through the test scenario acquisition channel of the data acquisition module, and simultaneously starting the channel adaptive expansion mechanism to dynamically increase the number of acquisition channels and adjust the channel acquisition parameters according to the number of devices under test and the type of scenario, thereby constructing a test environment that matches the test requirements.

[0014] Specifically, the process of obtaining multi-dimensional original matrix data is as follows: based on the type of the device under test and the preset basic dimensions of the test target, related dimensions are automatically derived through a dynamic dimension derivation mechanism; a three-dimensional anchor code with timestamps, spatial coordinates and device identifiers is assigned to the collected data points using spatiotemporal anchoring coding; and a matrix dynamic reconstruction algorithm is used to automatically fill the matrix with multi-source data according to the hierarchical structure of device units, time slices and environmental parameters to obtain a multi-dimensional original matrix with correlation.

[0015] Specifically, obtaining the corrected data matrix includes verifying the integrity of the device environment data of the multi-dimensional original matrix row by row according to the time series, and marking missing values ​​and outliers; for the missing values, an adaptive weighted interpolation algorithm based on data type is used to complete the continuous data and discrete data.

[0016] The anomaly root cause determination model distinguishes anomaly types. For outliers that need to retain trends, a trend constraint interpolation algorithm is used to incorporate the overall change trend of the data column when completing the data. The completed equipment environment data is integrated with the verified data to replace the missing and abnormal data in the multi-dimensional original matrix data, resulting in corrected matrix data whose data trend meets the needs of the scenario.

[0017] Specifically, the process of obtaining time-series unified data through data preprocessing includes: using a time-series alignment calibration engine to use channels whose sampling frequency matches the target time-series precision as the reference time series; using a dynamic interpolation frame-filling algorithm to supplement the missing timestamp data under the reference time series; using an adaptive downsampling algorithm to filter time point data reflecting data change characteristics from redundant timestamp data to obtain a time-series unified initial dataset; activating a data noise filtering model and filtering noise through a wavelet threshold denoising algorithm to obtain time-series unified standardized data.

[0018] Specifically, the structured storage process includes: assigning unique identifiers for test scenarios to time-series unified data using a unique data encoding algorithm, and establishing a mapping relationship between the identifiers and data content; employing a multi-dimensional data partitioning model to divide the data into independent storage partitions according to data type, and performing secondary classification within each partition using data credibility scoring; selecting a time-series optimized database to write the time-series unified data in timestamp order, while simultaneously enabling a real-time data backup mechanism, using an incremental data synchronization algorithm to synchronize the main database and the backup database in real time, constructing a data retrieval acceleration model, establishing a multi-dimensional indexing system, and performing structured storage for the time-series unified standardized data.

[0019] Specifically, obtaining the priority-ranked list of compatible devices includes: verifying the protocol compatibility of candidate devices based on device protocol standards, compatibility rules, and basic device attribute data stored in the database's built-in protocol model library; filtering a set of basic compatible devices with matching protocol types; and, based on the set of compatible devices and multi-dimensional environmental detection features, dynamically calculating feature weights according to features and scene priority using a weighted multi-feature collaborative filtering algorithm, calculating the similarity between the basic compatible devices and the scene requirement model, and ranking the basic compatible devices to obtain the priority-ranked list of compatible devices.

[0020] Specifically, the construction of the communication connection link between the device and the system includes: extracting protocol attribute data of the device to be adapted from the device resource library, retrieving basic communication data based on the test system, and passing the device unique identifier and system requirement identifier of the basic data to the dynamic decision tree input layer of the test protocol adaptation.

[0021] The protocol type matching node compares the device protocol type with the system supported protocols to generate judgment result data; the judgment result data is transmitted with the device attribute data to be verified to form a hierarchical data transmission link; the leaf node generates communication adaptation path data and adaptation risk warning data between the device and the system based on the judgment results of the previous level, and transmits the data to the system's communication link.

[0022] Specifically, the initialization configuration process of the real-time monitoring test equipment includes: using the test environment initialization configuration parameters extracted by the environment configuration module of the linked flexible test system and the real-time access status data of the equipment, after standardization processing, as input variables of the Bayesian inference model; calculating the posterior probability distribution of equipment access failures using the Bayesian formula based on the input variables and a preset historical fault prior probability library; evaluating the posterior probability distribution using preset fault threshold judgment rules to obtain the initialization configuration anomaly monitoring results and early warning information.

[0023] Specifically, the process of multi-linkage resource scheduling is as follows: a dynamic greedy and backtracking correction fusion algorithm is adopted. The test tasks are allocated accordingly through a greedy strategy to obtain an initial resource scheduling scheme. Based on the scheduling scheme, a backtracking correction mechanism is introduced to reallocate resources using the device resource elasticity coefficient. At the same time, the weight factor of resource scheduling is dynamically updated in combination with the real-time linkage requirements of the test scenario. Dynamic resource scheduling is carried out through iterative optimization.

[0024] Specifically, the process of matching the built-in protocol parsing component required for access testing with the communication environment is as follows: the component required by the test requirement protocol gene is located through the gene matching algorithm; the feature parameters are transformed into protocol parameter tuning factors using the dynamic mapping model between environmental features and protocol parameters; the protocol parsing component runs according to the tuned parameters and automatically adjusts the protocol parameters in reverse, forming a closed loop of protocol call, environment adaptation and verification optimization.

[0025] Specifically, the process of mapping and matching the discrete test data with the test requirement indicators includes: allocating basic weights according to the priority of the test scenario through a dynamic weight iteration algorithm, statistically analyzing the deviation between the discrete data and the requirement indicators, and automatically adjusting the weights of the corresponding data dimensions; and using a deviation attribution mapping module to associate the deviation values ​​between the data and the requirements with specific data dimensions to generate a mapping result with attribution.

[0026] The beneficial effects of this invention are as follows: relying on adaptive channel expansion, priority device lists, and dynamic greedy and backtracking scheduling technologies, the platform achieves unified integration and intelligent allocation of test resources. On the one hand, it breaks down the barriers of device dispersion, allowing idle devices to quickly match test scenarios through protocol compatibility verification and dynamic weight scheduling, significantly improving device sharing rates. On the other hand, adaptive expansion of acquisition channels and one-click generation of test systems eliminate the need for repeated manual setup and debugging, shortening the test solution delivery cycle, quickly responding to diverse customer needs, and reducing enterprise testing costs and resource waste.

[0027] Leveraging dynamic dimension derivation, adaptive interpolation completion, and unified temporal preprocessing technologies, the platform overcomes data quality challenges. It ensures the correlation of multi-source data through spatiotemporal anchoring encoding, employs type-adaptive interpolation and trend constraint algorithms to complete missing and anomalous data, and then performs temporal alignment and denoising processing to achieve microsecond-level data synchronization, resolving the issues of data asynchrony and incompleteness in traditional testing. Simultaneously, multi-dimensional partitioned structured storage ensures data security and traceability, providing a highly reliable data foundation for test result analysis and meeting customers' core demands for testing accuracy.

[0028] The platform utilizes protocol-adaptive dynamic decision tree and protocol gene mapping closed-loop technology, enabling rapid access for devices using multiple protocols such as CAN / CANFD and Modbus. It adapts to the testing needs of various scenarios, including multi-functional controllers for new energy vehicles and charging piles. Combined with dynamic weighted iterative mapping technology, it achieves intelligent matching of discrete data with required indicators, automatically generating attribution-based analysis results and replacing tedious manual operations. Simultaneously, Bayesian fault reasoning provides early warning of device initialization anomalies. Coupled with the platform's remote monitoring and intelligent alarm functions, it reduces testing safety risks, promotes the transformation of testing processes from fixed modes to flexible intelligence, and expands the scope of business coverage. Attached Figure Description

[0029] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0030] Figure 1 This is a schematic diagram of a linkage-type flexible testing system according to the present invention;

[0031] Figure 2 This is a block diagram of the linkage scheduling module in this invention; Detailed Implementation

[0032] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0033] Please see Figure 1-2 A linkage flexible testing system includes: a data acquisition module, a data processing module, a multi-linkage resource scheduling module, a flexible testing scheme enabling module, and an intelligent control and testing report module.

[0034] The data acquisition module obtains the corresponding test environment based on the linked flexible test system, collects data from the device under test, and acquires multi-dimensional raw matrix data;

[0035] The data processing module verifies the multi-dimensional original matrix data through data integrity monitoring, triggers the value compensation function when the data is abnormal, and obtains the corrected data matrix; it performs data preprocessing on the corrected data matrix to obtain time-series unified data, and stores the data in a structured manner.

[0036] The multi-linkage resource scheduling module obtains a priority-ranked list of compatible devices and the test devices required for the test scenario based on the structured storage of the data. Based on the test devices, it constructs a communication connection link between the devices and the system through a protocol adaptation dynamic decision tree algorithm, triggers a preset fault reasoning model, monitors the initialization configuration of the test devices in real time, and performs multi-linkage resource scheduling through a device resource scheduling algorithm.

[0037] The flexible testing solution enabling module utilizes the multi-linkage resource scheduling to preset the execution order, condition judgment rules, and loop parameters of the test steps; at the same time, it connects to the built-in protocol parsing component required for testing and matches it with the communication environment to generate an automated script for the flexible testing solution.

[0038] The intelligent control and testing report module captures data in real time during the testing process based on the automated script; through preset data association rules, it maps and matches the captured discrete test data with test requirement indicators to generate multi-dimensional test results.

[0039] Specifically, obtaining a test environment of the appropriate scale includes activating basic acquisition devices through the test scenario acquisition channels of the data acquisition module based on the number of devices under test, the type of test scenario, the density of test data acquisition, and test requirements. At the same time, an adaptive channel expansion mechanism is started to dynamically increase the number of acquisition channels and adjust the channel acquisition parameters according to the number of devices under test and the type of scenario, thereby constructing a test environment that matches the test requirements.

[0040] Specifically, the process of obtaining multi-dimensional original matrix data is as follows: based on the type of the device under test and the preset basic dimensions of the test target, related dimensions are automatically derived through a dynamic dimension derivation mechanism; a three-dimensional anchor code with timestamps, spatial coordinates and device identifiers is assigned to the collected data points using spatiotemporal anchoring coding; and a matrix dynamic reconstruction algorithm is used to automatically fill the matrix with multi-source data according to the hierarchical structure of device units, time slices and environmental parameters to obtain a multi-dimensional original matrix with correlation.

[0041] Specifically, obtaining the corrected data matrix includes verifying the integrity of the device environment data of the multi-dimensional original matrix row by row according to the time series, and marking missing values ​​and outliers; for the missing values, an adaptive weighted interpolation algorithm based on data type is used to complete the continuous data and discrete data.

[0042] The anomaly root cause determination model distinguishes anomaly types. For outliers that need to retain trends, a trend constraint interpolation algorithm is used to incorporate the overall change trend of the data column when completing the data. The completed equipment environment data is integrated with the verified data to replace the missing and abnormal data in the multi-dimensional original matrix data, resulting in corrected matrix data whose data trend meets the needs of the scenario.

[0043] In this embodiment, the testing requirements focus on the durability verification of the all-in-one controller under the coordinated operation of charging, power supply and driving. It is necessary to connect to various devices such as controller, battery simulator, power grid simulation source, motor load, environmental box, etc. Core electrical parameters need to be collected at high frequency to capture transient changes, and environmental parameters need to be dynamically monitored to reflect the impact of operating conditions.

[0044] Dynamic calculation formula for the number of acquisition channels:

[0045]

[0046] Where, N ch N represents the final number of acquisition channels, k1 is the device number influence coefficient (positively correlated with the parameter acquisition dimension of the device under test, with a larger value for multi-module devices), and N represents the number of acquisition channels. dev T represents the total number of devices under test, and T is the scene complexity correction factor (a larger value is taken for collaborative scenes due to increased module interaction). scnThe complexity level is quantified by module interaction logic. The number of channels can be automatically expanded based on the number of devices and the complexity of the scenario to ensure coverage of the core parameter collection needs of all devices and avoid data omissions due to insufficient channels.

[0047] Based on the multi-functional controller device type and collaborative durability test objectives, the basic dimensions of the device, module, and core parameters (such as controller, OBC, output voltage; controller, motor, three-phase current) are preset. Whether a related dimension is derived is determined by the mutual information value, as shown in the following formula:

[0048]

[0049] Where I(X,Y) is the mutual information value between the basic dimension X and the latent dimension Y (the larger the value, the stronger the correlation), p(x,y) is the joint probability distribution of X and Y, and p(x) and p(y) are the marginal probability distributions of X and Y, respectively. When I(X,Y)... I th (I th When the threshold is set to the correlation threshold, new dimensions are automatically derived.

[0050] The original matrix is ​​validated using time series analysis, and missing and outlier values ​​are marked. For continuous data (such as voltage and current), adaptive weighted interpolation is used for completion, as shown in the following formula:

[0051]

[0052] in, To complete the value, w j The weights (decreasing with time distance |ij| to ensure that nearest neighbor data has a greater impact), d j For adjacent normal data, n is the amount of neighboring data (dynamically adjusted according to the degree of data fluctuation) to ensure the continuity and rationality of the supplementary data.

[0053] For outliers that need to retain trends, trend-constrained interpolation is used to ensure that the completed data conforms to the physical changes in the scenario and avoids deviating from the overall trend. The completed data is then integrated with the validated normal data to replace the outliers and missing data in the original matrix, resulting in a corrected matrix whose data trend meets the requirements of the collaborative testing scenario. This provides a high-quality data foundation for subsequent time-series unification, structured storage, and test result analysis.

[0054] Specifically, the process of obtaining time-series unified data through data preprocessing includes: using a time-series alignment calibration engine to use channels whose sampling frequency matches the target time-series precision as the reference time series; using a dynamic interpolation frame-filling algorithm to supplement the missing timestamp data under the reference time series; using an adaptive downsampling algorithm to filter time point data reflecting data change characteristics from redundant timestamp data to obtain a time-series unified initial dataset; activating a data noise filtering model and filtering noise through a wavelet threshold denoising algorithm to obtain time-series unified standardized data.

[0055] In this embodiment, the formula for filtering out electromagnetic interference and noise generated by sensor jitter in the data is as follows:

[0056]

[0057] in, Denoising wavelet coefficients. W j,k These are the original wavelet coefficients (j is the wavelet decomposition scale, and k is the translation coefficient). Dynamic threshold (calculated from noise intensity) N is the data length. (Noise standard deviation);

[0058] Specifically, the structured storage process includes: assigning unique identifiers for test scenarios to time-series unified data using a unique data encoding algorithm, and establishing a mapping relationship between the identifiers and data content; employing a multi-dimensional data partitioning model to divide the data into independent storage partitions according to data type, and performing secondary classification within each partition using data credibility scoring; selecting a time-series optimized database to write the time-series unified data in timestamp order, while simultaneously enabling a real-time data backup mechanism, using an incremental data synchronization algorithm to synchronize the main database and the backup database in real time, constructing a data retrieval acceleration model, establishing a multi-dimensional indexing system, and performing structured storage for the time-series unified standardized data.

[0059] In this embodiment, during the structured storage stage, a unique code generation algorithm is first used to assign a unique code containing the test scenario type, test batch, and device identifier to standardized data, establishing a one-to-one mapping relationship between the code and the data content, ensuring that data can be quickly located and traced through the code. Then, a multi-dimensional data partitioning model is adopted to divide the data into independent storage partitions according to data type (such as electrical parameters, environmental parameters, and status parameters). Within each partition, secondary classification is performed based on the reliability score at the time of data acquisition (calculated based on the accuracy of the acquisition device and the stability of the link), facilitating the subsequent priority retrieval of high-reliability data for analysis. A time-series optimized database is selected for storage, writing standardized data in timestamp order, while simultaneously enabling a real-time backup mechanism. An incremental data synchronization algorithm synchronizes data between the main database and the backup database in real time to avoid data loss. Furthermore, a data retrieval acceleration model is constructed, establishing a multi-dimensional indexing system based on device identifier, parameter type, and time range, significantly improving the data retrieval efficiency in subsequent resource scheduling and result analysis stages.

[0060] Specifically, obtaining the priority-ranked list of compatible devices includes: verifying the protocol compatibility of candidate devices based on device protocol standards, compatibility rules, and basic device attribute data stored in the database's built-in protocol model library; filtering a set of basic compatible devices with matching protocol types; and, based on the set of compatible devices and multi-dimensional environmental detection features, dynamically calculating feature weights according to features and scene priority using a weighted multi-feature collaborative filtering algorithm, calculating the similarity between the basic compatible devices and the scene requirement model, and ranking the basic compatible devices to obtain the priority-ranked list of compatible devices.

[0061] Specifically, the construction of the communication connection link between the device and the system includes: extracting protocol attribute data of the device to be adapted from the device resource library, retrieving basic communication data based on the test system, and passing the device unique identifier and system requirement identifier of the basic data to the dynamic decision tree input layer of the test protocol adaptation.

[0062] The protocol type matching node compares the device protocol type with the system supported protocols to generate judgment result data; the judgment result data is transmitted with the device attribute data to be verified to form a hierarchical data transmission link; the leaf node generates communication adaptation path data and adaptation risk warning data between the device and the system based on the judgment results of the previous level, and transmits the data to the system's communication link.

[0063] In this embodiment, based on the basic set of compatible devices and the multi-dimensional environmental detection features of the current test scenario (such as the spatial layout of the test site, the stability of device power supply, and the network communication environment), the device compatibility is calculated using a weighted multi-feature collaborative filtering algorithm. Weights are dynamically assigned to different features (such as device response speed, operational stability, and physical distance from other devices) according to the priority of the test scenario. Among them, the device response speed, which is directly related to test accuracy, has the highest weight, and the physical distance, which is related to resource scheduling efficiency, has the second highest weight. Then, based on the weights, the similarity between each basic compatible device and the scenario requirement model is calculated, and the devices are sorted from high to low similarity to generate a priority-ranked list of compatible devices, ensuring that the device that best meets the scenario requirements is called first.

[0064] When constructing the communication connection link between the device and the system, protocol attribute data such as protocol version, data transmission format, and interface type of the device to be adapted is first extracted from the device resource library. At the same time, basic communication data such as communication port configuration and data interaction protocol standards are retrieved from the test system. The unique device identifier and system requirement identifier are then passed to the input layer of the dynamic decision tree for test protocol adaptation. The first-level node of the dynamic decision tree is the protocol type matching node, which compares the device protocol type with the protocol types supported by the system and generates three judgment results: matching, basic matching requiring conversion, and non-matching. If the judgment result is basic matching requiring conversion, it enters the next-level protocol conversion node, which determines the intermediate protocol and parameters required for conversion based on the device attribute data (such as interface parameters and data transmission rate). Finally, the leaf nodes generate specific communication adaptation paths (including direct connection paths or connection paths via protocol conversion modules) based on the judgment results of all previous levels, and output adaptation risk warnings (such as slight delays that may be caused by protocol conversion, connection nodes that need to be monitored, etc.). The adaptation path data is then passed to the system communication link to complete the communication connection between the device and the system.

[0065] Specifically, the initialization configuration process of the real-time monitoring test equipment includes: using the test environment initialization configuration parameters extracted by the environment configuration module of the linked flexible test system and the real-time access status data of the equipment, after standardization processing, as input variables of the Bayesian inference model; calculating the posterior probability distribution of equipment access failures using the Bayesian formula based on the input variables and a preset historical fault prior probability library; evaluating the posterior probability distribution using preset fault threshold judgment rules to obtain the initialization configuration anomaly monitoring results and early warning information.

[0066] In this embodiment, during the initial configuration of the monitoring and testing equipment, the initial configuration parameters corresponding to the test scenario (such as equipment rated operating parameters, communication port initialization settings, data acquisition threshold settings, etc.) are first extracted from the environment configuration module of the linked flexible testing system. Simultaneously, real-time status data of each device after connection (such as actual operating parameters, port connection status, data transmission initialization results, etc.) are collected. These two types of data are standardized and then input into the Bayesian inference model. The model, combined with a pre-set historical fault prior probability database (recording the probability of fault occurrence corresponding to different configuration deviations in past tests), calculates the posterior probability distribution of various faults under the current device connection status using the Bayesian formula. Then, it evaluates the posterior probability distribution using preset fault threshold judgment rules. If the posterior probability of a certain type of fault exceeds the threshold, it is determined to be an abnormal initial configuration, and an early warning message containing the fault type, possible causes, and troubleshooting suggestions is generated. If the posterior probability of all faults is lower than the threshold, the initial configuration is determined to be normal, allowing entry into the formal testing phase. The formula for calculating the device connection fault probability based on Bayesian inference is as follows:

[0067]

[0068] P(F|X): Posterior probability of fault F occurring given device state data X; P(X|F): Likelihood probability of observing state data X when fault F occurs (based on historical fault data statistics); P(F): Prior probability of fault F (set based on past test failure rates); P(X): Marginal probability of state data X (P(X) = P(X|F)). P(F)+P(X| F) P( F)), P( F) represents the fault-free prior probability.

[0069] Specifically, the process of multi-linkage resource scheduling is as follows: a dynamic greedy and backtracking correction fusion algorithm is adopted. The test tasks are allocated accordingly through a greedy strategy to obtain an initial resource scheduling scheme. Based on the scheduling scheme, a backtracking correction mechanism is introduced to reallocate resources using the device resource elasticity coefficient. At the same time, the weight factor of resource scheduling is dynamically updated in combination with the real-time linkage requirements of the test scenario. Dynamic resource scheduling is carried out through iterative optimization.

[0070] In this embodiment, dynamic greedy and backtracking resource scheduling optimization adjusts resource allocation based on device load and elasticity coefficient, and the formula is as follows:

[0071]

[0072] Among them, R new (D): Optimized resource allocation for device D; Rold (D): Initial resource allocation for device D; (D): Resource elasticity coefficient of equipment D (reflects performance stability when load changes); : The average elasticity coefficient of all devices participating in the scheduling; : Linkage demand coefficient.

[0073] Specifically, the process of matching the built-in protocol parsing component required for access testing with the communication environment is as follows: the component required by the test requirement protocol gene is located through the gene matching algorithm; the feature parameters are transformed into protocol parameter tuning factors using the dynamic mapping model between environmental features and protocol parameters; the protocol parsing component runs according to the tuned parameters and automatically adjusts the protocol parameters in reverse, forming a closed loop of protocol call, environment adaptation and verification optimization.

[0074] Specifically, the process of mapping and matching the discrete test data with the test requirement indicators includes: allocating basic weights according to the priority of the test scenario through a dynamic weight iteration algorithm, statistically analyzing the deviation between the discrete data and the requirement indicators, and automatically adjusting the weights of the corresponding data dimensions; and using a deviation attribution mapping module to associate the deviation values ​​between the data and the requirements with specific data dimensions to generate a mapping result with attribution.

[0075] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A linkage-type flexible testing system, characterized in that, include: Data acquisition module, data processing module, multi-linkage resource scheduling module, flexible testing solution enabling module, intelligent control and testing report module; The data acquisition module obtains the corresponding test environment based on the linked flexible test system, collects data from the device under test, and acquires multi-dimensional raw matrix data; The data processing module verifies the multi-dimensional original matrix data through data integrity monitoring, and triggers the compensation function to obtain the corrected data matrix when the data is abnormal. The corrected data matrix is ​​preprocessed to obtain time-series unified data, and the data is stored in a structured manner. The multi-linkage resource scheduling module obtains a priority-ranked list of compatible devices and the test devices required for the test scenario based on the structured storage of the data. Based on the test equipment, a communication link between the device and the system is constructed through a protocol-adaptive dynamic decision tree algorithm. At the same time, a preset fault reasoning model is triggered to monitor the initial configuration of the test equipment in real time, and multi-linkage resource scheduling is performed through a device resource scheduling algorithm. The flexible testing solution enabling module utilizes the multi-linkage resource scheduling to preset the execution order, condition judgment rules, and loop parameters of the test steps; at the same time, it connects to the built-in protocol parsing component required for testing and matches it with the communication environment to generate an automated script for the flexible testing solution. The intelligent control and testing report module captures data in real time during the testing process based on the automated script; By using preset data association rules, the captured discrete test data is mapped and matched with test requirement indicators to generate multi-dimensional test results.

2. The system according to claim 1, characterized in that, The process of obtaining the corresponding test environment includes activating the basic acquisition device through the test scenario acquisition channel of the data acquisition module based on the number of devices under test, the type of test scenario, the density of test data acquisition, and the test requirements. At the same time, the channel adaptive expansion mechanism is started to dynamically increase the number of acquisition channels and adjust the channel acquisition parameters according to the number of devices under test and the type of scenario, thereby constructing a test environment that matches the test requirements.

3. The system according to claim 1, characterized in that, The specific process of obtaining multi-dimensional raw matrix data is as follows: based on the type of the device under test and the test target, a basic dimension is preset, and related dimensions are automatically derived through a dynamic dimension derivation mechanism; a three-dimensional anchor code with timestamps, spatial coordinates and device identifiers is assigned to the collected data points using spatiotemporal anchoring coding. The matrix dynamic reconstruction algorithm automatically fills the matrix with multi-source data according to the hierarchical structure of device units, time slices and environmental parameters, and obtains a multi-dimensional original matrix with correlation.

4. The system according to claim 1, characterized in that, The process of obtaining the corrected data matrix includes: verifying the integrity of the device environment data of the multi-dimensional original matrix row by row according to the time series, and marking missing values; for the missing values ​​and outliers, using a data type adaptive weighted interpolation algorithm to complete the continuous data and discrete data. The anomaly root cause determination model distinguishes anomaly types. For outliers that need to retain trends, a trend constraint interpolation algorithm is used to incorporate the overall change trend of the data column when completing the data. The completed equipment environment data is integrated with the verified data to replace the missing and abnormal data in the multi-dimensional original matrix data, resulting in corrected matrix data whose data trend meets the needs of the scenario.

5. The system according to claim 1, characterized in that, The process of obtaining time-series unified data through data preprocessing includes: using a time-series alignment and calibration engine to use channels whose sampling frequency matches the target time-series precision as the reference time series; using a dynamic interpolation frame-filling algorithm to supplement the missing timestamp data under the reference time series; using an adaptive downsampling algorithm to filter time point data reflecting data change characteristics from redundant timestamp data to obtain a time-series unified initial dataset; activating a data noise filtering model and filtering noise through a wavelet threshold denoising algorithm to obtain time-series unified standardized data.

6. The system according to claim 1, characterized in that, The specific process of performing structured storage includes assigning a unique identifier for the test scenario to the time-series unified data through a data unique encoding generation algorithm, and establishing a mapping relationship between the identifier and the data content; A multi-dimensional data partitioning model is adopted, dividing the data into independent storage partitions according to data type. Within each partition, secondary classification is performed using data credibility scoring. A time-series optimized database is selected, and time-series unified data is written in timestamp order. At the same time, a real-time data backup mechanism is enabled, and incremental data synchronization algorithms are used to synchronize the main database and the backup database in real time. A data retrieval acceleration model is constructed, and a multi-dimensional indexing system is established to perform structured storage of the time-series unified standardized data.

7. The system according to claim 1, characterized in that, The process involves obtaining a priority-sorted list of compatible devices. This includes: verifying the protocol compatibility of candidate devices based on device protocol standards, compatibility rules, and basic device attribute data stored in the database's built-in protocol model library; filtering a set of basic compatible devices with matching protocol types; and, based on the set of compatible devices and multi-dimensional environmental detection features, dynamically calculating feature weights according to features and scene priority using a weighted multi-feature collaborative filtering algorithm, calculating the similarity between the basic compatible devices and the scene requirement model, sorting the basic compatible devices, and obtaining a priority-ranked list of compatible devices.

8. The system according to claim 1, characterized in that, The construction of the communication connection link between the device and the system includes: extracting protocol attribute data of the device to be adapted from the device resource library, retrieving basic communication data based on the test system, and passing the device unique identifier and system requirement identifier of the basic data to the dynamic decision tree input layer of the test protocol adaptation. The protocol type matching node compares the device protocol type with the system supported protocols to generate judgment result data; the judgment result data is transmitted with the device attribute data to be verified to form a hierarchical data transmission link; the leaf node generates communication adaptation path data and adaptation risk warning data between the device and the system based on the judgment results of the previous level, and transmits the data to the system's communication link.

9. The system according to claim 1, characterized in that, The specific process of initializing the real-time monitoring and testing equipment includes: using the test environment initialization configuration parameters extracted by the environment configuration module of the linked flexible testing system and the real-time access status data of the equipment, after standardization processing, as input variables of the Bayesian inference model; calculating the posterior probability distribution of equipment access failures using the Bayesian formula based on the input variables and a preset historical fault prior probability library; evaluating the posterior probability distribution using preset fault threshold judgment rules to obtain the initialization configuration anomaly monitoring results and early warning information.

10. The system according to claim 1, characterized in that, The specific process of multi-linkage resource scheduling is as follows: a dynamic greedy and backtracking correction fusion algorithm is adopted. The test tasks are allocated accordingly through the greedy strategy to obtain an initial resource scheduling scheme. Based on the scheduling scheme, a backtracking correction mechanism is introduced to reallocate resources using the device resource elasticity coefficient. At the same time, the weight factor of resource scheduling is dynamically updated in combination with the real-time linkage requirements of the test scenario. Dynamic resource scheduling is carried out through iterative optimization.

11. The system according to claim 1, characterized in that, The specific process of matching the built-in protocol parsing component required for access testing with the communication environment is as follows: the component required by the test requirement protocol gene is located by using the gene matching algorithm; the feature parameters are transformed into protocol parameter tuning factors by using the dynamic mapping model between environmental features and protocol parameters; the protocol parsing component runs according to the tuned parameters and automatically adjusts the protocol parameters in reverse, forming a closed loop of protocol call, environment adaptation and verification optimization.

12. The system according to claim 1, characterized in that, The specific process of mapping and matching discrete test data with test requirement indicators includes: allocating basic weights according to the priority of test scenarios through a dynamic weight iteration algorithm, statistically analyzing the deviation between discrete data and requirement indicators, and automatically adjusting the weights of corresponding data dimensions; and using a deviation attribution mapping module to associate the deviation values ​​between the data and requirements with specific data dimensions to generate attribution-based mapping results.

Images