Microcomputer system performance test and evaluation method, device, equipment and medium

Abstract

The invention relates to a microcomputer system performance test and evaluation method, device and equipment and a medium, and the microcomputer system performance test and evaluation method comprises the steps: if it is detected that a key event of a microcomputer system has occurred, executing a corresponding test cycle operation according to the key event; and converting and generating uniformly formatted real-time log data based on each test result in the test cycle operation, clustering and matching the real-time log data to a predefined abnormal state, and matching corresponding adjustment parameters based on the abnormal state to obtain the power state of the microcomputer system. Executing a first dynamic adjustment operation of the power state based on the performance adjustment parameter; according to the method, standardized real-time log data is generated from each test result, and the log data can be intelligently clustered and matched to a predefined abnormal state, so that the abnormity can be quickly identified, the adjustment parameters are generated, the power state and the test process of the system are dynamically optimized, and the flexibility of abnormity response and power consumption adjustment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of computer performance testing, and in particular to a microcomputer system performance testing and evaluation method, device, equipment and medium. Background Art

[0002] At present, microcomputer systems are widely used in industrial control, intelligent devices and embedded systems, and their performance testing is an important part of ensuring system stability and reliability. Performance testing usually requires simulating multiple workload scenarios, including high-load operation, low-power mode switching, and abnormal state detection. Existing microcomputer system performance testing methods usually record the system operation status and generate relevant test results through fixed test cycle operations. Such methods often rely on preset test parameters and fixed test scenarios, and have weak response to abnormal states and dynamic adjustment capabilities of power consumption; in addition, the analysis of log data mainly relies on manual intervention or simple rules, and lacks the support of intelligent abnormal analysis models, resulting in the inability to adapt to system state changes in a timely manner, thereby reducing test efficiency and accuracy. Summary of the invention

[0003] In order to solve the problem that the existing microcomputer system performance testing methods lack dynamic adjustment capabilities, resulting in the inability to adapt to system status changes in a timely manner, thereby reducing testing efficiency and accuracy, the present application provides a microcomputer system performance testing and evaluation method, device, equipment and medium.

[0004] The above-mentioned invention objective of the present application is achieved through the following technical solutions: A microcomputer system performance test and evaluation method, the microcomputer system performance test and evaluation method comprising: If it is detected that a critical event of the microcomputer system has occurred, a corresponding test cycle operation is performed according to the critical event; Based on each test result in the test cycle operation, convert and generate uniformly formatted real-time log data, the real-time log data including log entries corresponding to each test result; Based on the established abnormal analysis model, real-time log data is clustered and matched to a predefined abnormal state, and based on the abnormal state, the corresponding adjustment parameters are matched, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment; Acquiring a power state of the microcomputer system, and performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter; It is determined whether the response condition is satisfied, and if satisfied, a test adjustment parameter is generated according to the response condition, wherein the test adjustment parameter is used to dynamically adjust a second dynamic adjustment operation of the test cycle operation.

[0005] By adopting the above technical scheme, by detecting the key events of the microcomputer system and executing the corresponding test cycle operation according to the key events, the test process can dynamically respond to the real-time state of the system, thereby improving the flexibility and adaptability of the test; by converting and generating uniformly formatted real-time log data based on each test result in the test cycle operation, standardized data support can be provided for subsequent analysis, thereby improving the efficiency and accuracy of data processing; by clustering and matching the real-time log data to predefined abnormal states based on the established abnormal analysis model, and matching the corresponding adjustment parameters, intelligent identification of abnormal states can be achieved, thereby improving the accuracy and adaptability of abnormal detection; by obtaining the power state of the microcomputer system and executing the first dynamic adjustment operation of the power state based on the performance adjustment parameters, the system power allocation can be optimized to avoid excessive power consumption or waste of resources, thereby improving the stability and operation efficiency of the system.

[0006] In a preferred example, the present application may be further configured as follows: the step of converting and generating uniformly formatted real-time log data based on each test result in the test cycle operation, wherein the real-time log data includes log entries corresponding to each test result, includes: Extracting key data related to each test result, wherein the key data includes test time, test event type, hardware status parameters, and test result description; Determine the matching field of the key data, and match the corresponding format field according to the matching field, wherein the format field includes a cycle time period, a cycle number, a classification level and a key level; Determine a predefined entry format, and fill the format fields into the pre-filled fields corresponding to the entry format in sequence to generate a log entry; A predefined arrangement format is determined, and each of the log entries is arranged to generate corresponding real-time log data.

[0007] By adopting the above technical solution, by extracting key data related to each test result, the comprehensiveness and accuracy of the log content can be ensured, thereby providing a reliable data basis for subsequent analysis; by determining the matching fields of the key data and matching the corresponding format fields according to the matching fields, the standardized processing of the data format can be achieved, thereby improving the compatibility and analysis efficiency of the log data; by determining the predefined entry format and filling the format fields in the pre-filled sections corresponding to the entry format in sequence to generate log entries, the log structure can be unified, thereby simplifying data management; By determining a predefined arrangement format and arranging each log entry to generate real-time log data, it is possible to provide time-series data support for the test results, thereby facilitating the location and analysis of anomalies.

[0008] In a preferred example, the present application may be further configured as follows: the step of clustering and matching the real-time log data to a predefined abnormal state based on the established abnormal analysis model includes: Based on the established abnormal analysis model, locate the classification level in each of the log entries, wherein if the classification levels of two or more of the log entries are consistent, then determine that the test results corresponding to the two or more log entries are consistent; Classifying each of the log entries into a repetition group and a characteristic group according to the classification level; The feature groups are matched to predefined abnormal states.

[0009] By adopting the above technical solution, by locating the classification segment in each log entry based on the established abnormal analysis model, the log data can be quickly classified, thereby improving the analysis efficiency of the abnormal state; by classifying the log entries into repeated groups and feature groups according to the classification segments, data redundancy can be reduced and key features can be extracted, thereby optimizing the accuracy of abnormal analysis; by matching the feature groups to predefined abnormal states, system abnormalities can be accurately identified, thereby providing a scientific basis for subsequent adjustment operations.

[0010] In a preferred example, the present application may be further configured as follows: the step of matching the feature group to a predefined abnormal state includes: Calling a state database, extracting state keywords in the state database, and a matching order of the state keywords; Based on the established anomaly analysis model, locate the key segments of each of the log entries in the feature group; According to the matching order, matching the state keywords corresponding to the key segments; A mapping relationship is determined according to the state keywords, and a corresponding abnormal state is determined.

[0011] By adopting the above technical solution, by calling the state database and extracting the state keywords and their matching order in the state database, the context information of the exception analysis can be enriched, thereby improving the accuracy of the exception matching; by locating the key segments of the log entries in the feature group based on the exception analysis model, the core information related to the abnormal state can be extracted, thereby optimizing the classification and matching process of the abnormal state; by matching the state keywords corresponding to the key segments according to the matching order, the scope of the abnormal state can be gradually narrowed, thereby improving the efficiency of abnormal state identification; by determining a good mapping relationship according to the state keywords and determining the corresponding abnormal state, the log entry can be accurately associated with the abnormal state, thereby providing a reliable basis for the adjustment operation.

[0012] In a preferred example, the present application may be further configured as follows: the step of obtaining the power state of the microcomputer system and performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter includes: Monitor the current power state parameters of the microcomputer system in real time through hardware sensors or power management modules, wherein the current power state parameters include current CPU power consumption, current GPU power consumption and current whole machine power consumption; Determine a test power consumption parameter corresponding to an abnormal state of the performance adjustment parameter, wherein the test power consumption parameter includes a test CPU power consumption, a test GPU power consumption, and a test whole machine power consumption; Determine whether the current power state parameter is greater than or equal to the test power consumption parameter; If so, performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter; If not, the current power state parameters are maintained.

[0013] By adopting the above technical solution, the current power state parameters of the microcomputer system can be monitored in real time through hardware sensors or power management modules, so that the real-time operating power of the system can be obtained, thereby providing basic data support for dynamic adjustment; by determining the test power consumption parameters of the abnormal state corresponding to the performance adjustment parameters, the target range of the power consumption adjustment can be clarified, thereby improving the pertinence and scientific nature of the adjustment operation; by judging whether the current power state parameters are greater than or equal to the test power consumption parameters, it is possible to dynamically judge whether the system power consumption is abnormal, thereby ensuring that the adjustment operation is triggered only when necessary; by executing the first dynamic adjustment operation of the power state based on the performance adjustment parameters, it is possible to optimize power allocation and avoid the impact of excessive power consumption on the system, thereby improving the energy efficiency and stability of the system.

[0014] In a preferred example, the present application may be further configured as follows: the step of determining whether the response condition is satisfied, and if satisfied, generating a test adjustment parameter according to the response condition, wherein the test adjustment parameter is used in the step of dynamically adjusting the second dynamic adjustment operation of the test cycle operation, including: Determining a response condition that has been triggered, the response condition comprising the current power state parameter being less than the test power consumption parameter, the log entries in the feature group reaching a preset entry threshold, and the completion time period of the test cycle operation exceeding a preset time period; Combine the triggered response conditions into corresponding response combinations; The test adjustment parameters corresponding to the response combination are matched, wherein the test adjustment parameters include a cycle interval adjustment parameter, a test task increase / decrease adjustment parameter, and a test step increase / decrease adjustment parameter.

[0015] By adopting the above technical solution, by determining the response conditions that have been triggered, including the current power state parameter being less than the test power consumption parameter, the log entries in the feature group reaching the preset entry threshold, and the completion time period of the test cycle operation exceeding the preset time period, the trigger conditions that need to be adjusted can be fully identified, thereby ensuring the timeliness and reliability of dynamic adjustment; by combining the triggered response conditions into corresponding response combinations, it is possible to realize combined judgment of complex conditions, thereby improving the accuracy of adjustment parameter generation; by matching the test adjustment parameters corresponding to the response combination, including the cycle interval adjustment parameters, the test task increase and decrease adjustment parameters, and the test step increase and decrease adjustment parameters, the test cycle operation can be flexibly adjusted, thereby optimizing test efficiency and resource utilization.

[0016] The second object of the invention is achieved by the following technical solutions: A microcomputer system performance test and evaluation device, the microcomputer system performance test and evaluation device comprising: An execution module, for executing a corresponding test cycle operation according to the key event if it is detected that a key event of the microcomputer system has occurred; A generating module, configured to convert and generate uniformly formatted real-time log data based on each test result in the test cycle operation, wherein the real-time log data includes a log entry corresponding to each test result; A matching module, used to cluster and match real-time log data to predefined abnormal states based on an established abnormal analysis model, and to match corresponding adjustment parameters based on the abnormal states, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment; an acquisition module, configured to acquire a power state of the microcomputer system and perform a first dynamic adjustment operation of the power state based on the performance adjustment parameter; The judgment module is used to judge whether the response condition is met, and if so, to generate a test adjustment parameter according to the response condition, wherein the test adjustment parameter is used to dynamically adjust the second dynamic adjustment operation of the test cycle operation.

[0017] The third objective of the present application is achieved through the following technical solutions: A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned microcomputer system performance test and evaluation method are implemented.

[0018] The fourth objective of the present application is achieved through the following technical solutions: A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned microcomputer system performance test and evaluation method are implemented.

[0019] In summary, the present application includes at least one of the following beneficial technical effects: This application solves the problems of fixed test cycles and lack of intelligent response capabilities in existing microcomputer system performance test methods by introducing dynamic test cycle operations driven by key events, unified formatted real-time log data generation, intelligent abnormal analysis models, and dynamic adjustment mechanisms. Specifically, when a key event (such as high load, low power consumption, or abnormal state) is detected, the system can trigger the corresponding test cycle operation and generate standardized real-time log data for each test result for subsequent analysis. Through an unsupervised learning model based on a clustering algorithm, the system can intelligently cluster log data and match it to a predefined abnormal state, thereby quickly identifying anomalies and generating adjustment parameters. Adjustment parameters include performance adjustment and test adjustment, which are used to dynamically optimize the power state and test process of the system, such as adjusting hardware power consumption or modifying test task configuration, so as to adapt to system state changes in real time. This solution not only improves the flexibility of abnormal response and power consumption adjustment, but also significantly improves test efficiency and accuracy through intelligent log analysis, effectively ensuring the stability and reliability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flow chart of a microcomputer system performance testing and evaluation method in one embodiment of the present application.

[0021] Figure 2 is a flowchart for implementing step S20 in a microcomputer system performance test and evaluation method in one embodiment of the present application; Figure 3 is a flowchart for implementing step S30 in a microcomputer system performance test and evaluation method in one embodiment of the present application; Figure 4 is a flowchart for implementing step S303 in a microcomputer system performance test and evaluation method in one embodiment of the present application; Figure 5 is a flowchart for implementing step S40 in a microcomputer system performance test and evaluation method in one embodiment of the present application; Figure 6 is a flowchart for implementing step S50 in a microcomputer system performance test and evaluation method in one embodiment of the present application; Figure 7 This is a principle block diagram of a microcomputer system performance test and evaluation device in one embodiment of the present application; Figure 8 It is a schematic diagram of a device in an embodiment of the present application. DETAILED DESCRIPTION

[0022] The present application is further described in detail below in conjunction with the accompanying drawings.

[0023] In one embodiment, if Figure 1 As shown, the present application discloses a microcomputer system performance test and evaluation method, which specifically includes the following steps: S10, if it is detected that a critical event of the microcomputer system has occurred, executing a corresponding test cycle operation according to the critical event; In this embodiment, if it is detected that a critical event of the microcomputer system has occurred, the corresponding test cycle operation is performed according to the critical event. Critical events include but are not limited to system startup, shutdown, blue screen, hardware temperature is too high, voltage abnormality, operation jam, and predetermined test trigger conditions. Specifically, the system built-in event detection module monitors the status of the microcomputer system in real time through a variety of hardware and software sensors. For example, by reading the operating system log, detect whether a blue screen event (such as blue screen code 0x0000007E) occurs; obtain the real-time temperature data of the CPU and GPU through the temperature sensor to determine whether it exceeds the safety range (such as 80°C); confirm whether the mainboard power supply is stable through the voltage monitoring module. When any critical event is detected, the test system will start the corresponding test cycle operation. For example, after the blue screen event occurs, the system will perform a startup cycle test to check whether the abnormality recurs; when the temperature abnormality is detected, the load reduction test will be started to evaluate the heat dissipation performance and load handling capacity.

[0024] S20, based on each test result in the test cycle operation, convert and generate uniformly formatted real-time log data, the real-time log data including a log entry corresponding to each test result; In this embodiment, based on each test result in the test cycle operation, uniformly formatted real-time log data is converted and generated, and the real-time log data includes log entries corresponding to each test result. Specifically, the test system records the results of each step of the test cycle, including the time point of test execution, event type, specific steps of the test, and result status. For example, in a high-load CPU test, the system extracts the start time and end time of the test, and records the change data of CPU usage and temperature. These log entries are structured and stored through predefined fields to ensure data format consistency and facilitate subsequent processing and analysis. At the same time, the system supports deduplication and sorting operations on log entries to ensure the integrity of the time series and reduce the impact of redundant data on storage.

[0025] S30, based on the established abnormal analysis model, clustering and matching the real-time log data to a predefined abnormal state, and matching the corresponding adjustment parameters based on the abnormal state, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment; In this embodiment, based on the established abnormal analysis model, the real-time log data is clustered and matched to a predefined abnormal state, and based on the abnormal state, the corresponding adjustment parameters are matched, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment. Specifically, the abnormal analysis model clusters the features in the real-time log data through an unsupervised learning clustering algorithm (such as K-means, DBSCAN). For example, all log entries with CPU temperatures exceeding 90°C are clustered into an "overheating abnormality" group, and log entries with abnormally increased GPU power consumption are clustered into a "power consumption overload" group. Subsequently, the model matches these clustering results with a predefined abnormal state database. For example, the "overheating abnormality" group corresponds to an abnormal state of heat dissipation problems, and the system generates adjustment parameters based on this match, such as limiting the maximum frequency of the CPU and GPU or triggering a fan acceleration strategy. For the "power consumption overload" group, the overall operating load of the device is reduced, or switched to a low power consumption mode.

[0026] S40, acquiring a power state of the microcomputer system, and performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter; In this embodiment, the power state of the microcomputer system is obtained, and the first dynamic adjustment operation of the power state is performed based on the performance adjustment parameters. Specifically, the power consumption data of the system, including CPU power consumption, GPU power consumption and whole machine power consumption, is collected in real time through the system hardware sensor or power management module. For example, the system reads the power consumption interface of the motherboard and obtains that the current CPU power value is 65W, the GPU power value is 120W, and the total power consumption is 185W. According to the performance adjustment parameters, for example, the goal of controlling the CPU power consumption below 70W, the CPU operating frequency is dynamically adjusted (such as reduced to 2.8GHz) or the GPU voltage value is reduced, so as to achieve reasonable power distribution. At the same time, when the system detects that the power consumption is low, for example, the whole machine power consumption is less than 100W, the system can switch to high-performance mode to complete computing-intensive tasks more efficiently, thereby achieving a dynamic balance between performance and power consumption.

[0027] S50, determining whether the response condition is satisfied, and if satisfied, generating a test adjustment parameter according to the response condition, the test adjustment parameter being used for dynamically adjusting the second dynamic adjustment operation of the test cycle operation.

[0028] In this embodiment, it is determined whether the response condition is met. If so, a test adjustment parameter is generated according to the response condition, and the test adjustment parameter is used to dynamically adjust the second dynamic adjustment operation of the test loop operation. Specifically, the response condition includes whether the current power state parameter of the system is lower than the test power consumption parameter, for example, the current power consumption of the whole machine is lower than 120W; whether the abnormal log entries in the feature log group reach the preset threshold, for example, the number of GPU abnormal logs exceeds 5; whether the completion time of the test loop operation exceeds the preset time period, for example, the test takes more than 15 minutes. When these conditions are met, the system will generate test adjustment parameters according to the specific response conditions. For example, when the power consumption is too low, the system can increase the test load intensity, such as running more multi-threaded computing tasks; when the number of abnormal logs is large, the system can suspend the test to protect the hardware; when the test times out, the system can adjust the interval time of the test cycle or skip some repeated test steps, thereby improving the overall test efficiency and reliability. The adjusted parameters will be fed back to the test system in real time to ensure the effectiveness and pertinence of the dynamic adjustment operation.

[0029] Specifically, the process of matching the corresponding adjustment parameters based on the abnormal state is achieved by comparing and mapping the features in the real-time log data with the predefined abnormal state database through the abnormal analysis model. First, the abnormal analysis model performs cluster analysis on the log data and classifies log entries with similar features into the same group. Then, the system generates corresponding abnormal state identifiers based on the clustering results, such as "overheating abnormality", "power overload", "frequency instability", etc. Next, by looking up the mapping rules in the abnormal state database, for example, the adjustment parameter corresponding to "overheating abnormality" is to reduce the operating frequency of the CPU and GPU, and the adjustment parameter corresponding to "power overload" is to switch to low power consumption mode, the system can quickly determine the adjustment parameters associated with the abnormal state. This matching process combines static predefined rules with dynamic real-time data analysis to ensure that the adjustment parameters can accurately make the best adjustment decisions for the current system state.

[0030] In summary, by detecting key events of the microcomputer system and executing corresponding test cycle operations according to the key events, the test process can dynamically respond to the real-time state of the system, thereby improving the flexibility and adaptability of the test; by converting and generating uniformly formatted real-time log data based on each test result in the test cycle operation, standardized data support can be provided for subsequent analysis, thereby improving the efficiency and accuracy of data processing; by clustering and matching real-time log data to predefined abnormal states based on the established abnormal analysis model, and matching corresponding adjustment parameters, intelligent identification of abnormal states can be achieved, thereby improving the accuracy and adaptability of abnormal detection; by obtaining the power state of the microcomputer system and executing the first dynamic adjustment operation of the power state based on the performance adjustment parameters, the system power allocation can be optimized to avoid excessive power consumption or waste of resources, thereby improving the stability and operation efficiency of the system; This application solves the problems of fixed test cycles and lack of intelligent response capabilities in existing microcomputer system performance test methods by introducing dynamic test cycle operations driven by key events, unified formatted real-time log data generation, intelligent abnormal analysis models, and dynamic adjustment mechanisms. Specifically, when a key event (such as high load, low power consumption, or abnormal state) is detected, the system can trigger the corresponding test cycle operation and generate standardized real-time log data for each test result for subsequent analysis. Through an unsupervised learning model based on a clustering algorithm, the system can intelligently cluster log data and match it to a predefined abnormal state, thereby quickly identifying anomalies and generating adjustment parameters. The adjustment parameters include performance adjustment and test adjustment, which are used to dynamically optimize the power state and test process of the system, such as adjusting hardware power consumption or modifying test task configuration, so as to adapt to system state changes in real time. This solution not only improves the flexibility of abnormal response and power consumption adjustment, but also significantly improves test efficiency and accuracy through intelligent log analysis, effectively ensuring the stability and reliability of the system.

[0031] In one embodiment, if Figure 2 As shown, in step S20, that is, based on each test result in the test cycle operation, converting and generating uniformly formatted real-time log data, the real-time log data includes a log entry corresponding to each test result, including: S201, extracting key data related to each test result, wherein the key data includes test time, test event type, hardware status parameters and test result description; In this embodiment, key data related to each test result is extracted, wherein the key data includes test time, test event type, hardware status parameters, and test result description. Specifically, test time refers to the time point when the test run starts and ends, for example, a CPU high-load test starts at 10:00:00 on January 14, 2025 and ends at 10:10:00; the test event type describes the test category currently being executed, such as startup test, load test, or power consumption test; hardware status parameters include real-time collected data such as CPU temperature, GPU power consumption, and motherboard voltage, such as CPU temperature 85°C, GPU power consumption 120W; the test result description records the execution status of the test, such as "success", "failure", or specific abnormal information, such as "power consumption overload". By extracting these key data, the comprehensiveness of the log content can be ensured, providing basic support for subsequent data processing and analysis.

[0032] S202, determining the matching fields of the key data, and matching the corresponding format fields according to the matching fields, wherein the format fields include a cycle time period, a cycle number, a classification level, and a key level; In this embodiment, the matching fields of the key data are determined, and the corresponding format fields are matched according to the matching fields. The format fields include the cycle time period, the number of cycles, the classification segment and the key segment. Specifically, the matching field refers to the standardized information extracted from the key data, such as the test time and the event type as the matching field, which are used to calculate the cycle time period and the number of cycles. The cycle time period refers to the time spent on a round of testing, for example, the time period generated by the test start time 10:00:00 and the end time 10:10:00 is 10 minutes; the number of cycles indicates the number of repeated runs of the current test, such as the third test cycle; the classification segment indicates the classification identification of the test results through the abnormal analysis model, such as "overheating" and "stable"; the key segment refers to a specific data segment identified in the test, such as "CPU load exceeds 95%". By matching these fields, the test data can be standardized, laying the foundation for the formatting of log entries.

[0033] S203, determining a predefined entry format, and filling the format fields into the pre-filled sections corresponding to the entry format in sequence to generate a log entry; In this embodiment, a predefined entry format is determined, and the format fields are sequentially filled into the pre-filled sections corresponding to the entry format to generate a log entry. Specifically, the predefined entry format refers to a log record template designed according to system requirements, for example, an entry contains fields: timestamp, event type, state parameters, and test results. The aforementioned format fields are filled into the predefined template to generate a standardized log entry. For example, the timestamp "2025-01-14 10:10:00", the event type "load test", the state parameter "CPU temperature: 85°C", and the test result "passed" constitute a complete entry: {"timestamp": "2025-01-14 10:10:00", "event_type": "load test", "state_param": "CPU temperature: 85°C", "result": "passed"}. This process ensures the consistency and structuring of log entries, which is convenient for subsequent storage and processing.

[0034] S204: Determine a predefined arrangement format, and arrange the log entries to generate corresponding real-time log data.

[0035] In this embodiment, a predefined arrangement format is determined, and each log entry is arranged to generate corresponding real-time log data. Specifically, the predefined arrangement format refers to a rule for sorting log entries by time or event type, such as arranging log entries in chronological order to intuitively present the test process and timeline. Assuming that there are 3 log entries, the corresponding timestamps are "2025-01-14 10:00:00", "2025-01-14 10:05:00" and "2025-01-14 10:10:00", the system will arrange these entries in chronological order to generate a set of real-time log data: [Entry 1, Entry 2, Entry 3]. Through this arrangement, the real-time log data can not only fully reflect the test status, but also provide time series support for abnormal location and problem tracking.

[0036] In summary, by extracting key data related to each test result, the comprehensiveness and accuracy of the log content can be ensured, thereby providing a reliable data basis for subsequent analysis; by determining the matching fields of key data and matching the corresponding format fields according to the matching fields, the standardized processing of the data format can be achieved, thereby improving the compatibility and analysis efficiency of the log data; by determining the predefined entry format and filling the format fields in the pre-filled sections corresponding to the entry format in sequence to generate log entries, the log structure can be unified, thereby simplifying data management; By determining a predefined arrangement format and arranging each log entry to generate real-time log data, it is possible to provide time-series data support for the test results, thereby facilitating the location and analysis of anomalies.

[0037] In one embodiment, if Figure 3 As shown, in step S30, that is, based on the established abnormal analysis model, the step of clustering and matching the real-time log data to the predefined abnormal state includes: S301, based on the established abnormal analysis model, locate the classification level in each log entry, wherein, if the classification levels of two or more log entries are consistent, then determine that the test results corresponding to the two or more log entries are consistent; In this embodiment, based on the established abnormal analysis model, the classification segment in each log entry is located, wherein, if the classification segments of two or more log entries are consistent, the test results corresponding to the two or more log entries are determined to be consistent. Specifically, the classification segment is a classification identifier generated by the abnormal analysis model based on the characteristic fields in the log entry, which is used to describe the abnormal state category to which the log entry belongs. For example, the log entry contains the characteristic fields "CPU temperature", "GPU power consumption" and "system response time", and the abnormal analysis model will classify it as "normal", "overheating" or "power consumption abnormality" according to the numerical range of these fields. When two or more log entries have the same classification segment, for example, they are all marked as "overheating", the system determines that the test results corresponding to these entries are consistent, indicating that the system has the same type of abnormalities in this stage. This method can quickly classify the test results, thereby improving the efficiency of abnormality detection.

[0038] S302, classifying each log entry into a repetition group and a characteristic group according to the classification level; In this embodiment, each log entry is classified into a repetition group and a feature group according to the classification level. Specifically, the repetition group refers to a set of log entries with the same classification level and the same content, which is used to identify repeated abnormal situations; the feature group refers to a set of log entries with the same classification level but different content, which is used to extract abnormal entries with similar characteristics. For example, if the content recorded in the log entries is "CPU temperature exceeds 85°C, test failed", these entries are classified into the same repetition group; and if the log entries all show "GPU power consumption exceeds 100W", but the specific power consumption values ​​are different (such as 120W and 150W), these entries are classified into the same feature group. Through this grouping method, the system can effectively reduce redundant log data, extract abnormal features, and provide accurate input for subsequent analysis.

[0039] S303: Match to a predefined abnormal state according to the feature group.

[0040] In this embodiment, the feature group is matched to a predefined abnormal state. Specifically, the log entry in the feature group determines the corresponding abnormal state by comparing its feature field with the definition rules in the abnormal state database. For example, the log entry in the feature group shows that the GPU power consumption exceeds 120W, and the system extracts the abnormal state "power consumption overload" corresponding to the matching rule "GPU power consumption exceeds 100W and the test fails" from the abnormal state database. After the matching is completed, the system will mark this feature group as a "power consumption overload" state and associate adjustment parameters, such as reducing the GPU operating frequency or switching to energy-saving mode. Through this matching method, the system can accurately map complex log data to abnormal states, thereby providing a scientific basis for subsequent dynamic adjustments.

[0041] In summary, by locating the classification segment in each log entry based on the established anomaly analysis model, the log data can be quickly classified, thereby improving the analysis efficiency of abnormal conditions; by classifying the log entries into repeated groups and feature groups according to the classification segments, data redundancy can be reduced and key features can be extracted, thereby optimizing the accuracy of anomaly analysis; by matching the feature groups to predefined abnormal conditions, system anomalies can be accurately identified, thereby providing a scientific basis for subsequent adjustment operations.

[0042] In one embodiment, if Figure 4 As shown, in step S303, i.e., the step of matching to a predefined abnormal state according to the feature group, it includes: S3031, calling the state database, extracting the state keywords in the state database, and the matching order of the state keywords; In this embodiment, the state database is called to extract state keywords and the matching order of the state keywords in the state database. Specifically, the state database is a storage module for the correspondence between a predefined abnormal state and its characteristic keywords and priorities. Each abnormal state contains a set of characteristic keywords, which describe the characteristics of the abnormality and define the matching order according to importance. For example, for the abnormal state "power consumption overload", its state keywords include "power consumption", "current" and "temperature", and the matching order is "power consumption > current > temperature", indicating that "power consumption" is the feature that needs to be matched with the highest priority, and "temperature" is used as an auxiliary feature. By calling the state database, the system can load these state keywords and their priority order, and provide clear rules for subsequent abnormal matching. This design can ensure that the most relevant features are highlighted in the keyword matching process, while reducing the interference of irrelevant features on the matching process, thereby improving the efficiency and accuracy of abnormal identification.

[0043] S3032. Based on the established abnormal analysis model, locate the key segments of each log entry in the feature group; In this embodiment, based on the established abnormal analysis model, the key segments of each log entry in the feature group are located. Specifically, the abnormal analysis model is an intelligent analysis tool trained using a clustering algorithm or a feature extraction algorithm, which is used to identify and extract data segments that are critical to the abnormal state from the log entry. The key segment refers to the field in the log entry that directly affects the judgment of the abnormal state, such as "GPU power consumption", "power supply current", "system response time", etc. The model will automatically identify and mark the key segments based on the feature distribution and historical data of the log entry. For example, for a group of log entries of a feature group, the "GPU power consumption" field is displayed as 120W, which is significantly higher than the normal range, while other fields (such as CPU temperature and fan speed) are within the normal range. The system will locate "GPU power consumption" as the key segment of this group. This method of automatically locating key segments can greatly reduce manual intervention, improve the accuracy and efficiency of feature extraction, and lay the foundation for subsequent keyword matching.

[0044] S3033, matching the state keywords corresponding to the key segments in the matching order; In this embodiment, the state keywords corresponding to the key segments are matched according to the matching order. Specifically, the system compares the data in the key segments with the definition rules of the keywords one by one according to the keyword matching order extracted from the state database. For example, for the "power consumption overload" state, the matching order is "power consumption > current > temperature". The system will first check whether the "GPU power consumption" of 120W in the key segment exceeds the normal range (for example, the defined 100W threshold); if the match is successful, the system determines that there is an abnormal state of "power consumption overload" in the feature group; if the match is not successful, the next keyword "current" will continue to be matched. This step-by-step matching method can effectively narrow the judgment range of the abnormal state, give priority to capturing the most abnormal information, and take into account the reference value of other secondary features. For example, if "power consumption" and "current" are matched successfully at the same time, but "temperature" is not matched successfully, the system can still determine that the abnormal state is "power consumption overload" based on the main features, thereby avoiding misjudgment caused by excessive reliance on secondary features.

[0045] S3034. Determine a good mapping relationship according to the state keyword and determine the corresponding abnormal state.

[0046] In this embodiment, the mapping relationship determined according to the state keywords is used to determine the corresponding abnormal state. Specifically, the mapping relationship of the state keywords is a set of rules stored in the state database, and each rule defines a corresponding relationship between a set of keyword combinations and abnormal states. For example, "GPU power consumption>100W, current>20A" is mapped to the "power overload" state, and "CPU temperature>85°C" is mapped to the "overheating abnormality" state. When the keyword matching is completed, the system searches for the corresponding abnormal state according to the keyword combination that successfully matches. For example, in the log entry of the feature group, if "GPU power consumption is 120W" and "current is 22A" are matched successfully at the same time, the system will determine that the abnormal state is "power overload" according to the mapping rule. This process can structurally associate complex log data with specific abnormal states, ensure the scientificity and accuracy of abnormal judgment, and provide a reliable basis for subsequent dynamic adjustment operations (such as reducing GPU frequency or switching to low power mode). At the same time, the scalability of this mapping mechanism allows the system to flexibly add new state keywords and mapping rules as business needs change, further enhancing the adaptability and intelligence of the system.

[0047] In summary, by calling the state database and extracting the state keywords and their matching order in the state database, the context information of the exception analysis can be enriched, thereby improving the accuracy of exception matching; by locating the key segments of the log entries in the feature group based on the exception analysis model, the core information related to the abnormal state can be extracted, thereby optimizing the classification and matching process of the abnormal state; by matching the state keywords corresponding to the key segments according to the matching order, the scope of the abnormal state can be gradually narrowed, thereby improving the efficiency of abnormal state identification; by determining a good mapping relationship based on the state keywords and determining the corresponding abnormal state, the log entry can be accurately associated with the abnormal state, thereby providing a reliable basis for adjustment operations.

[0048] In one embodiment, if Figure 5 As shown, in step S40, the power state of the microcomputer system is obtained, and the step of performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter includes: S401, monitoring the current power state parameters of the microcomputer system in real time through a hardware sensor or a power management module, wherein the current power state parameters include the current CPU power consumption, the current GPU power consumption and the current whole machine power consumption; In this embodiment, the current power state parameters of the microcomputer system are monitored in real time through a hardware sensor or a power management module, wherein the current power state parameters include the current CPU power consumption, the current GPU power consumption and the current whole machine power consumption. Specifically, the hardware sensor can be integrated inside the motherboard or CPU / GPU, and obtains real-time data such as power, current, and voltage through a standard interface (such as PMBus, SMBus, or I2C). For example, the CPU sensor can provide power consumption data sampled per second, the GPU monitoring module can accurately record the dynamic power consumption of the graphics card, and the whole machine power consumption can be measured by an external power meter or power supply monitoring module. Assuming that the current CPU power consumption is 65W, the GPU power consumption is 120W, and the total power consumption is 185W, these data are transmitted to the power management module in real time to form a complete system operation power state diagram. The significance of real-time monitoring is to ensure that the system can accurately perceive the current power situation, provide high-reliability basic data for subsequent adjustment operations, and avoid abnormal power consumption out of control due to delayed monitoring.

[0049] S402, determining a test power consumption parameter corresponding to an abnormal state of the performance adjustment parameter, where the test power consumption parameter includes a test CPU power consumption, a test GPU power consumption, and a test whole machine power consumption; In this embodiment, the test power consumption parameters corresponding to the abnormal state of the performance adjustment parameters are determined, and the test power consumption parameters include test CPU power consumption, test GPU power consumption and test whole machine power consumption. Specifically, the performance adjustment parameters are generated by the abnormal analysis model according to the log analysis results and the abnormal state, and different power consumption target ranges are set for different abnormal states. For example, for the "abnormal heat dissipation" state, the system sets the test power consumption parameters to a CPU power consumption of no more than 50W, a GPU power consumption of no more than 90W, and a total power consumption of no more than 140W to reduce heat generation; and for the "power overload" state, the test power consumption parameters are set to a CPU power consumption of less than 60W, a GPU power consumption of less than 100W, and a total power consumption of no more than 160W. The setting of these parameters takes into account the hardware design limitations and system load requirements, and can effectively guide subsequent dynamic power adjustment operations, so that the system can meet performance requirements while keeping power within a safe range.

[0050] S403, determining whether the current power state parameter is greater than or equal to the test power consumption parameter; In this embodiment, it is determined whether the current power state parameter is greater than or equal to the test power consumption parameter. Specifically, the system compares the current power state parameter monitored in real time with the test power consumption parameter item by item to determine whether there is a power consumption exceeding the limit. For example, if the current CPU power consumption is 65W, the GPU power consumption is 120W, and the total power consumption is 185W, and the test power consumption parameters respectively stipulate that the CPU power consumption should be less than 60W, the GPU power consumption should be less than 100W, and the total power consumption should be less than 160W, the system will detect that the current state exceeds the parameter range and mark it as "power consumption overload". This judgment mechanism can quickly identify power consumption anomalies, especially under high load or abnormal working environment, and provide a trigger signal for subsequent dynamic adjustment. Through this real-time judgment method, the system can efficiently respond to power anomalies and avoid hardware damage or system instability caused by excessive power.

[0051] S404: If yes, perform a first dynamic adjustment operation of the power state based on the performance adjustment parameter; In this embodiment, if so, a first dynamic adjustment operation of the power state is performed based on the performance adjustment parameters. Specifically, the system dynamically adjusts the excess power consumption according to the performance adjustment parameters. For example, when the GPU power consumption exceeds the test power consumption parameter of 120W, the system can reduce power consumption by reducing the operating frequency of the GPU (such as from 1.5GHz to 1.2GHz) or reducing the voltage (such as from 1.2V to 1.0V). Similarly, for the case where the CPU power consumption exceeds the limit, the system can enable energy-saving mode and limit its operation to a lower frequency range (such as from 3.2GHz to 2.6GHz). In addition, in the case of overload of the whole machine power consumption, the system alleviates the overall load by limiting the operating power consumption of certain non-essential modules (such as network cards or external devices). Through these adjustment operations, the system can control the power consumption within a safe range while meeting performance requirements, thereby ensuring hardware security and system stability.

[0052] S405: If not, maintain the current power state parameters.

[0053] In this embodiment, if not, the current power state parameters are maintained. Specifically, when the system determines that the current power state parameters do not exceed the test power consumption parameter range, the power consumption adjustment operation will not be triggered, and the system will maintain the current operating mode. For example, when the CPU power consumption is 55W, the GPU power consumption is 95W, and the total power consumption is 150W, all within the test power consumption parameter range, the system will maintain the existing power allocation and operating configuration. The effect of maintaining the current power state parameters is to reduce unnecessary adjustment operations, ensure the continuity and efficiency of system performance, and avoid the potential impact of frequent adjustments on hardware. Through this mechanism, the system can achieve dynamic balance according to the actual operating state, achieve optimal adaptation between power consumption and performance, and make the overall operation more stable and efficient.

[0054] In summary, by real-time monitoring of the current power state parameters of the microcomputer system through hardware sensors or power management modules, the real-time operating power of the system can be obtained, thereby providing basic data support for dynamic adjustment; by determining the test power consumption parameters of the abnormal state corresponding to the performance adjustment parameters, the target range of the power consumption adjustment can be clarified, thereby improving the pertinence and scientific nature of the adjustment operation; by judging whether the current power state parameters are greater than or equal to the test power consumption parameters, it is possible to dynamically judge whether the system power consumption is abnormal, thereby ensuring that the adjustment operation is triggered only when necessary; by executing the first dynamic adjustment operation of the power state based on the performance adjustment parameters, it is possible to optimize power allocation and avoid the impact of excessive power consumption on the system, thereby improving the energy efficiency and stability of the system.

[0055] In one embodiment, if Figure 6 As shown, in step S50, it is determined whether the response condition is met. If so, a test adjustment parameter is generated according to the response condition. The test adjustment parameter is used in the second dynamic adjustment operation step of dynamically adjusting the test cycle operation, including: S501, determining a response condition that has been triggered, the response condition including that a current power state parameter is less than a test power consumption parameter, a log entry in a feature group reaches a preset entry threshold, and a completion time period of a test cycle operation exceeds a preset time period; In this embodiment, the triggered response conditions are determined, and the response conditions include that the current power state parameter is less than the test power consumption parameter, the log entry in the feature group reaches the preset entry threshold, and the completion time period of the test cycle operation exceeds the preset time period. Specifically, the current power state parameter is less than the test power consumption parameter, which means that the power data monitored by the system in real time is lower than the preset target range. For example, the test power consumption parameter stipulates that the CPU power consumption should be above 50W and the GPU power consumption should be above 80W, while the current actual power consumption is 45W and 75W respectively. This state indicates that the system is under low load or insufficient testing; the log entry in the feature group reaches the preset entry threshold, which means that the number of abnormal entries in a set of feature logs exceeds the set value, for example, the log entry of the feature group contains more than 10 records of "abnormal CPU temperature fluctuation", indicating that there are continuous problems in the operation of the system; the completion time period of the test cycle operation exceeds the preset time period, which means that the execution time of the test task exceeds the set reasonable range, for example, the test cycle is required to be completed within 10 minutes, and the current test has been running for 15 minutes but has not yet ended. This situation implies that the test efficiency is reduced or a jamming problem occurs. These response conditions can accurately capture the key factors that affect the test results by real-time monitoring and analysis of system status, providing a basis for subsequent adjustments.

[0056] S502, combining the triggered response conditions into a corresponding response combination; In this embodiment, the triggered response conditions are combined into corresponding response combinations. Specifically, the response combination is a logical set composed of multiple triggered response conditions, which is used to describe the comprehensive characteristics of the current state of the system. For example, when the system detects that the two response conditions of "the current power state parameter is lower than the test power consumption parameter" and "the abnormal log entry in the feature group reaches the threshold" are triggered at the same time, the system will combine them into a response combination of "low power consumption and frequent abnormalities"; if the condition of "test cycle timeout" is triggered at the same time, a response combination of "low power consumption, frequent abnormalities and test timeout" is formed. This response combination can be constructed by logical rules, such as AND or OR relationships. By combining response conditions into response combinations, the system can process complex state information more efficiently and provide accurate judgment basis for subsequent adjustment operations.

[0057] S503 , matching and responding to the test adjustment parameters corresponding to the combination, the test adjustment parameters including a cycle interval adjustment parameter, a test task increase / decrease adjustment parameter, and a test step increase / decrease adjustment parameter.

[0058] In this embodiment, the test adjustment parameters corresponding to the response combination are matched, and the test adjustment parameters include a cycle interval adjustment parameter, a test task increase and decrease adjustment parameter, and a test step increase and decrease adjustment parameter. Specifically, the system maps the response combination to the corresponding adjustment strategy by searching the adjustment parameter database. For example, for the response combination of "low power consumption and frequent abnormalities", the adjustment parameters matched by the system are: extending the cycle interval to reduce the system load (cycle interval adjustment parameter), increasing the test task to cover more load scenarios (test task increase and decrease adjustment parameter); for the response combination of "test timeout and frequent abnormalities", the adjustment parameters matched by the system are: reducing low-priority test steps (test step increase and decrease adjustment parameter) to shorten the test time and focus on key issues. The matching process combines the characteristics of the response combination and the adaptability of the adjustment parameters, optimizes the test process through precise adjustment strategies, and improves the test efficiency and the accuracy of the results. For example, in actual applications, if the power consumption of the GPU fluctuates frequently and the test takes too long, the system can quickly improve the test effect by reducing the test task load and optimizing the step sequence, while ensuring the test coverage and system stability. In this way, the test system can dynamically adapt and precisely adjust to ensure efficient operation, with customized settings based on actual conditions.

[0059] To summarize, by determining the response conditions that have been triggered, including the current power state parameter being less than the test power consumption parameter, the log entries in the feature group reaching the preset entry threshold, and the completion time period of the test loop operation exceeding the preset time period, the trigger conditions that need to be adjusted can be fully identified, thereby ensuring the timeliness and reliability of dynamic adjustment; by combining the triggered response conditions into corresponding response combinations, the combined judgment of complex conditions can be achieved, thereby improving the accuracy of adjustment parameter generation; by matching the test adjustment parameters corresponding to the response combination, including the loop interval adjustment parameters, the test task increase and decrease adjustment parameters, and the test step increase and decrease adjustment parameters, the test loop operation can be flexibly adjusted, thereby optimizing test efficiency and resource utilization.

[0060] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0061] In one embodiment, a microcomputer system performance test and evaluation device is provided, and the microcomputer system performance test and evaluation device corresponds one-to-one to a microcomputer system performance test and evaluation method in the above embodiment. Figure 7 As shown, the microcomputer system performance test and evaluation device includes an execution module, a generation module, a matching module, an acquisition module and a judgment module. The detailed description of each functional module is as follows: A microcomputer system performance test and evaluation device, the microcomputer system performance test and evaluation device comprising: An execution module, for executing a corresponding test cycle operation according to the key event if it is detected that a key event of the microcomputer system has occurred; A generating module, configured to convert and generate uniformly formatted real-time log data based on each test result in the test cycle operation, wherein the real-time log data includes a log entry corresponding to each test result; A matching module, used to cluster and match real-time log data to predefined abnormal states based on an established abnormal analysis model, and to match corresponding adjustment parameters based on the abnormal states, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment; an acquisition module, configured to acquire a power state of the microcomputer system and perform a first dynamic adjustment operation of the power state based on the performance adjustment parameter; The judgment module is used to judge whether the response condition is met, and if so, to generate a test adjustment parameter according to the response condition, wherein the test adjustment parameter is used to dynamically adjust the second dynamic adjustment operation of the test cycle operation.

[0062] Optionally, the generating module includes: An extraction unit, used to extract key data related to each test result, wherein the key data includes test time, test event type, hardware status parameters and test result description; A first determining unit is used to determine a matching field of the key data, and match a corresponding format field according to the matching field, wherein the format field includes a cycle time period, a cycle number, a classification level, and a key level; A second determining unit, configured to determine a predefined entry format, and sequentially fill the format fields into pre-filled sections corresponding to the entry format to generate a log entry; A third determining unit, configured to determine a predefined arrangement format, and arrange each of the log entries to generate corresponding real-time log data; Optionally, the matching module includes: a positioning unit, configured to locate the classification level in each of the log entries based on the established abnormality analysis model, wherein if the classification levels of two or more of the log entries are consistent, then it is determined that the test results corresponding to the two or more log entries are consistent; A classification unit, used for classifying each of the log entries into a repetition group and a characteristic group according to the classification level; A first matching unit, configured to match the feature group to a predefined abnormal state; Optionally, the matching unit includes: An extraction subunit, used for calling a state database, extracting state keywords in the state database, and a matching order of the state keywords; A positioning subunit, used for locating the key segments of each of the log entries in the feature group based on the established abnormality analysis model; A matching subunit, used for matching the state keywords corresponding to the key segments according to the matching order; A determination subunit, used to determine a good mapping relationship according to the state keyword, and determine a corresponding abnormal state; Optionally, the acquisition module includes: A monitoring unit, used to monitor the current power state parameters of the microcomputer system in real time through a hardware sensor or a power management module, wherein the current power state parameters include current CPU power consumption, current GPU power consumption and current whole machine power consumption; A fourth determining unit, configured to determine a test power consumption parameter corresponding to an abnormal state of the performance adjustment parameter, wherein the test power consumption parameter includes a test CPU power consumption, a test GPU power consumption, and a test whole machine power consumption; A judging unit, used to judge whether the current power state parameter is greater than or equal to the test power consumption parameter; an execution unit, configured to, if yes, perform a first dynamic adjustment operation of the power state based on the performance adjustment parameter; a maintaining unit, for maintaining the current power state parameters if no; Optionally, the judging module includes: a fifth determining unit, configured to determine a response condition that has been triggered, the response condition comprising that the current power state parameter is less than the test power consumption parameter, the log entries in the feature group reach a preset entry threshold, and the completion time period of the test cycle operation exceeds a preset time period; A combination unit, used to combine the triggered response conditions into corresponding response combinations; The second matching unit is used to match the test adjustment parameters corresponding to the response combination, wherein the test adjustment parameters include a cycle interval adjustment parameter, a test task increase / decrease adjustment parameter, and a test step increase / decrease adjustment parameter.

[0063] For the specific definition of a microcomputer system performance test and evaluation device, please refer to the definition of a microcomputer system performance test and evaluation method mentioned above, which will not be repeated here. Each module in the above-mentioned microcomputer system performance test and evaluation device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0064] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 8 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a microcomputer system performance test and evaluation method is implemented.

[0065] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: S10, if it is detected that a critical event of the microcomputer system has occurred, executing a corresponding test cycle operation according to the critical event; S20, based on each test result in the test cycle operation, convert and generate uniformly formatted real-time log data, the real-time log data including a log entry corresponding to each test result; S30, based on the established abnormal analysis model, clustering and matching the real-time log data to a predefined abnormal state, and matching the corresponding adjustment parameters based on the abnormal state, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment; S40, acquiring a power state of the microcomputer system, and performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter; S50, determining whether the response condition is satisfied, and if satisfied, generating a test adjustment parameter according to the response condition, the test adjustment parameter being used for dynamically adjusting the second dynamic adjustment operation of the test cycle operation.

[0066] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: S10, if it is detected that a critical event of the microcomputer system has occurred, executing a corresponding test cycle operation according to the critical event; S20, based on each test result in the test cycle operation, convert and generate uniformly formatted real-time log data, the real-time log data including a log entry corresponding to each test result; S30, based on the established abnormal analysis model, clustering and matching the real-time log data to a predefined abnormal state, and matching the corresponding adjustment parameters based on the abnormal state, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment; S40, acquiring a power state of the microcomputer system, and performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter; S50, determining whether the response condition is satisfied, and if satisfied, generating a test adjustment parameter according to the response condition, the test adjustment parameter being used for dynamically adjusting the second dynamic adjustment operation of the test cycle operation.

[0067] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0068] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0069] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A microcomputer system performance test and evaluation method, characterized in that: The microcomputer system performance test and evaluation method comprises: If it is detected that a critical event of the microcomputer system has occurred, a corresponding test cycle operation is performed according to the critical event; Based on each test result in the test cycle operation, convert and generate uniformly formatted real-time log data, the real-time log data including log entries corresponding to each test result; Based on the established abnormal analysis model, real-time log data is clustered and matched to a predefined abnormal state, and based on the abnormal state, the corresponding adjustment parameters are matched, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment; Acquiring a power state of the microcomputer system, and performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter; It is determined whether the response condition is satisfied, and if satisfied, a test adjustment parameter is generated according to the response condition, wherein the test adjustment parameter is used to dynamically adjust a second dynamic adjustment operation of the test cycle operation.

2. A microcomputer system performance test and evaluation method according to claim 1, characterized in that: The step of converting and generating uniformly formatted real-time log data based on each test result in the test cycle operation, wherein the real-time log data includes a log entry corresponding to each test result, comprises: Extracting key data related to each test result, wherein the key data includes test time, test event type, hardware status parameters, and test result description; Determine the matching field of the key data, and match the corresponding format field according to the matching field, wherein the format field includes a cycle time period, a cycle number, a classification level and a key level; Determine a predefined entry format, and fill the format fields into the pre-filled fields corresponding to the entry format in sequence to generate a log entry; A predefined arrangement format is determined, and each of the log entries is arranged to generate corresponding real-time log data.

3. A microcomputer system performance test and evaluation method according to claim 2, characterized in that: The step of clustering and matching the real-time log data to the predefined abnormal state based on the established abnormal analysis model includes: Based on the established abnormal analysis model, locate the classification level in each of the log entries, wherein if the classification levels of two or more of the log entries are consistent, then determine that the test results corresponding to the two or more log entries are consistent; Classifying each of the log entries into a repetition group and a characteristic group according to the classification level; The feature groups are matched to predefined abnormal states.

4. A microcomputer system performance test and evaluation method according to claim 3, characterized in that: The step of matching the feature group to a predefined abnormal state includes: Calling a state database, extracting state keywords in the state database, and a matching order of the state keywords; Based on the established anomaly analysis model, locate the key segments of each of the log entries in the feature group; According to the matching order, matching the state keywords corresponding to the key segments; A mapping relationship is determined according to the state keywords, and a corresponding abnormal state is determined.

5. A microcomputer system performance test and evaluation method according to claim 3, characterized in that: The step of obtaining the power state of the microcomputer system and performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter comprises: Monitor the current power state parameters of the microcomputer system in real time through hardware sensors or power management modules, wherein the current power state parameters include current CPU power consumption, current GPU power consumption and current whole machine power consumption; Determine a test power consumption parameter corresponding to an abnormal state of the performance adjustment parameter, wherein the test power consumption parameter includes a test CPU power consumption, a test GPU power consumption, and a test whole machine power consumption; Determine whether the current power state parameter is greater than or equal to the test power consumption parameter; If so, performing a first dynamic adjustment operation of the power state based on the performance adjustment parameter; If not, the current power state parameters are maintained.

6. A microcomputer system performance test and evaluation method according to claim 5, characterized in that: The step of determining whether the response condition is satisfied, and if satisfied, generating a test adjustment parameter according to the response condition, wherein the test adjustment parameter is used to dynamically adjust the second dynamic adjustment operation of the test cycle operation, includes: Determining a response condition that has been triggered, the response condition comprising the current power state parameter being less than the test power consumption parameter, the log entries in the feature group reaching a preset entry threshold, and the completion time period of the test cycle operation exceeding a preset time period; Combine the triggered response conditions into corresponding response combinations; The test adjustment parameters corresponding to the response combination are matched, wherein the test adjustment parameters include a cycle interval adjustment parameter, a test task increase / decrease adjustment parameter, and a test step increase / decrease adjustment parameter.

7. A microcomputer system performance test and evaluation device, characterized in that: The microcomputer system performance test and evaluation device comprises: An execution module, for executing a corresponding test cycle operation according to the key event if it is detected that a key event of the microcomputer system has occurred; A generating module, configured to convert and generate uniformly formatted real-time log data based on each test result in the test cycle operation, wherein the real-time log data includes a log entry corresponding to each test result; A matching module, used to cluster and match real-time log data to predefined abnormal states based on an established abnormal analysis model, and to match corresponding adjustment parameters based on the abnormal states, wherein the abnormal analysis model is an unsupervised learning model based on a clustering algorithm, and the adjustment parameters include performance adjustment parameters and response conditions for test adjustment; an acquisition module, configured to acquire a power state of the microcomputer system and perform a first dynamic adjustment operation of the power state based on the performance adjustment parameter; The judgment module is used to judge whether the response condition is met, and if so, to generate a test adjustment parameter according to the response condition, wherein the test adjustment parameter is used to dynamically adjust the second dynamic adjustment operation of the test cycle operation.

8. A microcomputer system performance test and evaluation device according to claim 7, characterized in that: The generation module comprises: An extraction unit, used to extract key data related to each test result, wherein the key data includes test time, test event type, hardware status parameters and test result description; A first determining unit is used to determine a matching field of the key data, and match a corresponding format field according to the matching field, wherein the format field includes a cycle time period, a cycle number, a classification level, and a key level; A second determining unit, configured to determine a predefined entry format, and sequentially fill the format fields into pre-filled sections corresponding to the entry format to generate a log entry; A third determining unit, configured to determine a predefined arrangement format, and arrange each of the log entries to generate corresponding real-time log data; The matching module comprises: a positioning unit, configured to locate the classification level in each of the log entries based on the established abnormality analysis model, wherein if the classification levels of two or more of the log entries are consistent, then it is determined that the test results corresponding to the two or more log entries are consistent; A classification unit, used for classifying each of the log entries into a repetition group and a characteristic group according to the classification level; A first matching unit, configured to match the feature group to a predefined abnormal state; The matching unit comprises: An extraction subunit, used for calling a state database, extracting state keywords in the state database, and a matching order of the state keywords; A positioning subunit, used for locating the key segments of each of the log entries in the feature group based on the established abnormality analysis model; A matching subunit, used for matching the state keywords corresponding to the key segments according to the matching order; A determination subunit, used to determine a good mapping relationship according to the state keyword, and determine a corresponding abnormal state; The acquisition module comprises: A monitoring unit, used to monitor the current power state parameters of the microcomputer system in real time through a hardware sensor or a power management module, wherein the current power state parameters include current CPU power consumption, current GPU power consumption and current whole machine power consumption; A fourth determining unit, configured to determine a test power consumption parameter corresponding to an abnormal state of the performance adjustment parameter, wherein the test power consumption parameter includes a test CPU power consumption, a test GPU power consumption, and a test whole machine power consumption; A judging unit, used to judge whether the current power state parameter is greater than or equal to the test power consumption parameter; an execution unit, configured to, if yes, perform a first dynamic adjustment operation of the power state based on the performance adjustment parameter; a maintaining unit, for maintaining the current power state parameters if no; The judging module comprises: a fifth determining unit, configured to determine a response condition that has been triggered, the response condition comprising that the current power state parameter is less than the test power consumption parameter, the log entries in the feature group reach a preset entry threshold, and the completion time period of the test cycle operation exceeds a preset time period; A combination unit, used to combine the triggered response conditions into corresponding response combinations; The second matching unit is used to match the test adjustment parameters corresponding to the response combination, wherein the test adjustment parameters include a cycle interval adjustment parameter, a test task increase / decrease adjustment parameter, and a test step increase / decrease adjustment parameter.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of a microcomputer system performance testing and evaluation method as described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a microcomputer system performance testing and evaluation method as described in any one of claims 1 to 6 are implemented.

Images