Adaptive generation method of high-speed interface fault-tolerant test process based on reinforcement learning

Abstract

The present invention relates to the technical field of high-speed interface fault-tolerant testing, specifically to a method for adaptively generating a high-speed interface fault-tolerant testing process based on reinforcement learning. First, high-speed interface data, environmental data, and historical fault-tolerant testing data are obtained; the fault-tolerant testing impact prediction module is trained using the historical fault-tolerant testing data to obtain a pre-training module and historical fault-tolerant testing impact weights; based on the high-speed interface data and environmental data, the state, action, strategy, and reward elements of the reinforcement learning agent are defined; and a high-speed interface adaptive test is performed in combination with the reinforcement learning agent, the pre-training module, and the test execution and monitoring platform. During the test process, the pre-training module is used to update the historical fault-tolerant testing impact weights and the current state; based on the test results and state changes obtained by the test execution and monitoring platform through monitoring the interface test state, error information, and performance indicators, an adaptive reward is obtained; the present invention improves the effectiveness and reliability of the high-speed interface fault-tolerant testing process.

Description

Technical Field

[0001] The present invention relates to the technical field of high-speed interface fault-tolerant testing, and in particular to a method for adaptively generating a high-speed interface fault-tolerant testing process based on reinforcement learning. Background Art

[0002] The overall performance and reliability of high-computing chips are highly dependent on the stable operation of their integrated high-speed interfaces. These interfaces not only transmit data at extremely high rates but also employ complex communication protocols with a vast range of configurable parameters. Furthermore, their proper operation is extremely sensitive to physical layer conditions and environmental factors.

[0003] During chip design verification and production test, one of the core challenges is to effectively and efficiently detect potential design flaws, manufacturing defects, or runtime functional errors in these high-speed interfaces. These errors can manifest in various forms, including but not limited to: bit errors in data transmission, cyclic redundancy check failures, protocol state machine errors, link training failures, packet loss or corruption, timing violations, and intermittent failures caused by signal integrity issues or environmental stress.

[0004] Existing high-speed interface testing and monitoring methods, such as specification-based or manually designed test cases, random or pseudo-random testing, and traditional coverage-driven verification, even with built-in error detection mechanisms, cannot adaptively adjust test strategies based on real-time test feedback (such as detected error type, frequency, or interface status changes). Furthermore, the lack of effective adaptive error detection mechanisms results in a significant consumption of test resources in low-risk areas, making it difficult to focus on deeply exploring the most problematic corner cases and failure boundaries. This directly impacts the efficiency and thoroughness of error detection.

[0005] To this end, a reinforcement learning-based adaptive generation method for high-speed interface fault-tolerant test process is proposed. Summary of the Invention

[0006] The purpose of the present invention is to provide a method for adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning. First, high-speed interface data, environmental data and historical fault-tolerant test data are obtained; the fault-tolerant test impact prediction module is trained using the historical fault-tolerant test data to obtain a pre-training module and historical fault-tolerant test impact weights; based on the high-speed interface data and environmental data, the state, action, strategy and reward elements of the reinforcement learning agent are defined; and a high-speed interface adaptive test is performed in combination with the reinforcement learning agent, the pre-training module and the test execution and monitoring platform. During the test process, the pre-training module is used to update the historical fault-tolerant test impact weights and the current state; based on the test results and state changes obtained by the test execution and monitoring platform through monitoring the interface test state, error information and performance indicators, an adaptive reward is obtained; the present invention improves the effectiveness and reliability of the high-speed interface fault-tolerant test process.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The adaptive generation method of high-speed interface fault-tolerant test process based on reinforcement learning includes:

[0009] Acquire high-speed interface data, environmental data, and historical fault-tolerance test data;

[0010] Using the historical fault-tolerant test data to train a fault-tolerant test impact prediction module to obtain a pre-trained fault-tolerant test impact prediction module and historical fault-tolerant test impact weights;

[0011] defining the state, action, strategy, and reward elements of a reinforcement learning agent based on the high-speed interface data and the environment data;

[0012] High-speed interface adaptive testing is performed in combination with the reinforcement learning agent, the pre-trained fault-tolerant test impact prediction module, and the test execution and monitoring platform. The testing process is as follows: setting the initial interface configuration and environmental conditions, and selecting the initial action based on the interface prior knowledge; obtaining the current state, and using the pre-trained fault-tolerant test impact prediction module to update the historical fault-tolerant test impact weight and the current state to obtain an updated state; selecting an action according to the current strategy and the updated state, and the test execution and monitoring platform executing the action instruction; obtaining the test results and the next state after executing the action based on the platform's monitoring of the interface test state, test error information, and test performance indicators; obtaining an adaptive reward based on the test results and state changes; updating the agent strategy using the experience tuple; and iterating through continuous interaction and learning with the environment until the test reaches a preset time, the number of errors found, the coverage target, or the agent strategy converges.

[0013] Furthermore, the high-speed interface data includes interface configuration parameter data and interface operation status data; the environmental data includes environmental parameter data and interference data; and the historical fault-tolerant test data includes historical test configuration data, historical test process data, and historical test result data.

[0014] Furthermore, the process of using the historical fault-tolerant test data to train the fault-tolerant test impact prediction module to obtain the pre-trained fault-tolerant test impact prediction module and the historical fault-tolerant test impact weights includes:

[0015] Constructing the fault-tolerance test impact prediction module;

[0016] Selecting historical test process data and historical test result data that match the configuration parameters of the interface to be tested from the historical fault-tolerant test data; wherein the historical test process data includes historical interface operation status data and historical environment data;

[0017] The historical test process data and the historical test result data are input into the fault-tolerant test impact prediction module for processing to obtain the pre-trained fault-tolerant test impact prediction module and the historical fault-tolerant test impact weight under the interface configuration to be tested.

[0018] Furthermore, the process of updating the historical fault-tolerant test impact weight and the current state using the pre-trained fault-tolerant test impact prediction module to obtain the updated state includes:

[0019] Get the current state;

[0020] Wherein, the current state includes the high-speed interface data and the environmental data;

[0021] Inputting the interface operation status data, environmental parameter data and interference data in the current state into the pre-trained fault-tolerant test impact prediction module for processing, and updating the historical fault-tolerant test impact weight to generate an impact prediction vector;

[0022] The impact prediction vector is combined with the current state to obtain the updated state.

[0023] Furthermore, based on the platform's monitoring of the interface test status, test error information, and test performance indicators, the test results and the next state after the action is executed are obtained, and the adaptive reward is obtained based on the test results and state changes.

[0024] The test execution and monitoring platform monitors the status, errors and performance of the high-speed interface after receiving the action instruction;

[0025] Comparing the interface test state obtained by monitoring with the next state to obtain a state distance reward;

[0026] Evaluate the severity and novelty of the monitored test error information and obtain error rewards;

[0027] Compare each test performance indicator obtained through monitoring with the performance benchmark value and obtain performance rewards;

[0028] A weighted sum is performed on the state distance reward, the error reward, and the performance reward to obtain the adaptive reward.

[0029] Furthermore, the experience tuple includes: the updated state, the action, the adaptive reward and the next state.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. The present invention proposes a high-speed interface fault-tolerant test process adaptive generation method based on deep learning to realize interface fault-tolerant adaptive testing; this method combines reinforcement learning, a fault-tolerant test impact prediction module, and a test execution and monitoring platform. The use of reinforcement learning and the test execution and monitoring platform can automatically execute the test process, reduce manual intervention, and shorten the test cycle; the use of the fault-tolerant test impact prediction module can assist the reinforcement learning agent in generating more effective test strategies.

[0032] 2. The present invention proposes a fault-tolerant test impact prediction module for generating an impact prediction vector; the module uses historical fault-tolerant test data for training to learn the correlation between the interface operating status and environmental parameters and the generation of fault-tolerant test errors; the generated impact prediction vector is used to update the current status in the high-speed interface adaptive test, thereby helping the reinforcement learning agent to generate an effective test strategy.

[0033] 3. The present invention proposes an adaptive reward to guide the agent to learn a better test strategy; this reward function combines the monitoring information of the test execution and monitoring platform with the agent status, and comprehensively considers the status, errors and performance indicators of the high-speed interface; this reward function is obtained by weighting the state distance reward, error reward and performance reward, and can evaluate the state exploration value, the severity and novelty of the error and the actual deviation of the performance, which is conducive to assisting the agent to learn a better test strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Schematic diagram of the process of adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning of the present invention;

[0035] Figure 2 Schematic diagram of the structure of the fault-tolerant test impact prediction module of the present invention;

[0036] Figure 3 It is a flowchart of the high-speed interface adaptive testing process of the present invention. DETAILED DESCRIPTION

[0037] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0038] See also Figures 1 to 3 The present invention provides a method for adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning. The technical solution is as follows:

[0039] Example 1:

[0040] In order to adaptively generate a high-speed interface fault-tolerant test process, a company used the high-speed interface fault-tolerant test process adaptive generation method based on reinforcement learning proposed in this invention. The process diagram of this method is as follows: Figure 1 As shown, specifically including:

[0041] Acquire high-speed interface data, environmental data, and historical fault-tolerance test data;

[0042] Furthermore, high-speed interface data, environmental data, and historical fault-tolerance test data can be obtained by calling the interface test management platform;

[0043] Furthermore, the high-speed interface data includes interface configuration parameter data and interface operation status data; the environmental data includes environmental parameter data and interference data; the historical fault-tolerant test data includes: historical test configuration data, historical test process data and historical test result data;

[0044] Furthermore, interface configuration parameter data includes: rate, bit width, pre-emphasis / equalization settings, timing parameters, etc.; interface operation status data includes: link training status, bit error rate (BER), cyclic redundancy check (CRC) error count, protocol error flag, throughput, etc.; environmental parameter data includes: core voltage, IO voltage, chip temperature, ambient temperature, etc.; interference data includes: jitter, noise, voltage drop, etc.

[0045] Furthermore, the historical test configuration data is historical interface configuration parameter data; the historical test process data includes historical interface operation status data and historical environment data; the historical test result data includes historical test normal result data and historical test abnormal result data;

[0046] Furthermore, the abnormal error types in the historical test abnormal result data include: CRC error, link retraining, link loss, BER exceeding the standard, etc.

[0047] By introducing high-speed interface data, environmental data and historical fault-tolerance test data, data support is provided for subsequent fault-tolerance test impact prediction and high-speed interface adaptive testing, thereby improving the effectiveness and reliability of the high-speed interface fault-tolerance test process.

[0048] The fault-tolerant test impact prediction module is trained using historical fault-tolerant test data to obtain a pre-trained fault-tolerant test impact prediction module and historical fault-tolerant test impact weights;

[0049] Furthermore, the process of training the fault-tolerance test impact prediction module using historical fault-tolerance test data to obtain the pre-trained fault-tolerance test impact prediction module and the historical fault-tolerance test impact weights includes:

[0050] Build a fault-tolerant test impact prediction module;

[0051] Selecting historical test process data and historical test result data that match the configuration parameters of the interface to be tested from the historical fault-tolerant test data; wherein the historical test process data includes historical interface operation status data and historical environment data;

[0052] The historical test process data and the historical test result data are input into the fault-tolerant test impact prediction module for processing to obtain the pre-trained fault-tolerant test impact prediction module and the historical fault-tolerant test impact weight under the interface configuration to be tested.

[0053] Furthermore, the fault-tolerant test impact prediction module is a time-environment interaction attention prediction network with the following structure: Figure 2 As shown, it includes: input layer, parallel temporal feature encoding layer, cross-modal interactive attention layer, feature fusion and compression layer, influence prediction head and influence weight generation layer;

[0054] Furthermore, the input layer is used to receive historical interface operation status data, historical environment data, and historical test result data; wherein the historical test result data is used as a label during module training;

[0055] Furthermore, a parallel temporal feature encoding layer is used to extract respective time-dependent features from the interface state data sequence and the environment state sequence; the parallel temporal feature encoding layer includes an interface state encoder and an environment state encoder; the interface state encoder and the environment state encoder are processed using a long short-term memory network and a temporal convolutional network, respectively, to obtain encoded interface state sequence features and environment state sequence features;

[0056] Furthermore, the cross-modal interaction attention layer is used to model the dynamic interaction between the interface state sequence and the environment state sequence. The cross-modal interaction attention layer adopts the cross-modal attention mechanism to calculate the attention weights of the environment on the interface state and the interface state on the environment, and obtains the weighted interface state features and environment state features respectively.

[0057] In this embodiment, the attention weight of the environment to the interface state is calculated by using the encoded interface state sequence feature as the query and the encoded environment state sequence feature as the key and value.

[0058] Furthermore, the feature fusion and compression layer is used to fuse the original encoding features and the attention-weighted interaction features into a fixed-length fused feature vector; the feature fusion and compression layer processing process can be expressed as:

[0059] ;

[0060] in, is the fusion feature vector; is a multi-layer perceptron; is the scratch weight value; is the encoded interface state sequence feature; is the encoded sequence feature of the environmental state; is the weighted interface status feature; is the weighted environmental state feature; The pooling operation in the time dimension is used to extract global context information;

[0061] Furthermore, the impact prediction head is used to output the final fault-tolerant test impact prediction value; this structure mainly uses a multi-layer perceptron and an activation function to process the fused feature vector to obtain the fault-tolerant test impact prediction value;

[0062] Furthermore, the influence weight generation layer is used to output the historical fault-tolerance test influence weight; the influence weight generation layer extracts the attention weights calculated in the cross-modal interaction attention layer and aggregates them along the time dimension to obtain the historical fault-tolerance test influence weight;

[0063] Furthermore, the fault-tolerant test impact prediction value and the fault-tolerant test impact weight constitute an impact prediction vector.

[0064] By utilizing the fault-tolerant test impact prediction module, the correlation between the interface operating status and environmental parameters and the occurrence of fault-tolerant test errors can be learned; and the generated impact prediction vector is used to update the current status in the high-speed interface adaptive test, helping the reinforcement learning agent to generate effective test strategies, thereby improving the effectiveness and reliability of the high-speed interface fault-tolerant test process.

[0065] Define the state, action, strategy, and reward elements of the reinforcement learning agent based on high-speed interface data and environmental data;

[0066] Furthermore, the status includes: interface configuration parameters, interface operating status, environmental parameters, interference data, and historical test information; wherein the historical test information includes the type of error found, the configuration / status area covered, etc.;

[0067] Further, the actions include: adjusting interference and environmental parameters, selecting test patterns or data traffic patterns, adjusting system load or operating status, etc.;

[0068] Furthermore, the strategy of the reinforcement learning agent is based on Q-value, that is, DQN is used as the reinforcement learning algorithm;

[0069] Furthermore, rewards include: state distance rewards, error rewards, and performance rewards.

[0070] Combine reinforcement learning agents, pre-trained fault-tolerant test impact prediction modules, and a test execution and monitoring platform for high-speed interface adaptive testing.

[0071] Furthermore, the process flow of the high-speed interface adaptive test process is shown as follows: Figure 3 As shown, specifically: set the initial interface configuration and environmental conditions, select the initial action based on the interface prior knowledge; obtain the current state, and use the pre-trained fault-tolerant test impact prediction module to update the historical fault-tolerant test impact weight and the current state to obtain the updated state; select the action according to the current strategy and the updated state, and the test execution and monitoring platform executes the action instruction; based on the platform's monitoring of the interface test state, test error information and test performance indicators, obtain the test results and the next state after the action is executed; obtain adaptive rewards based on the test results and state changes; use the experience tuple to update the agent strategy; and iterate through continuous interaction and learning with the environment until the test reaches the preset time, the number of errors found, the coverage target, or the agent strategy converges.

[0072] Furthermore, the interface prior knowledge includes: historical experience data, interface protocol specifications, and physical and environmental characteristics knowledge; this prior knowledge can be obtained through interface technical documents and historical test reports;

[0073] By combining reinforcement learning, the fault-tolerant test impact prediction module, and the test execution and monitoring platform, the reinforcement learning and test execution and monitoring platform can automatically execute the test process, reduce manual intervention, and shorten the test cycle; the fault-tolerant test impact prediction module can assist the reinforcement learning agent in generating more effective test strategies, thereby improving the effectiveness and reliability of the high-speed interface fault-tolerant test process.

[0074] Furthermore, the pre-trained fault-tolerant test impact prediction module is used to update the historical fault-tolerant test impact weights and the current state. The process of obtaining the updated state includes:

[0075] Get the current state;

[0076] The current status includes high-speed interface data and environmental data;

[0077] Input the current interface operation status data, environmental parameter data, and interference data into the pre-trained fault tolerance test impact prediction module for processing, and update the historical fault tolerance test impact weights to generate an impact prediction vector;

[0078] The impact prediction vector is combined with the current state to obtain the updated state.

[0079] By introducing the impact prediction vector obtained by the fault-tolerant test impact prediction module into the current state, the state information can be given predictive, explanatory and context-aware capabilities based on historical experience, highlighting the key influencing factors in the current state, improving the intelligence level of subsequent decision-making, and significantly improving test efficiency and defect detection capabilities, thereby ensuring the effectiveness and reliability of the high-speed interface fault-tolerant test process.

[0080] Furthermore, based on the platform's monitoring of the interface test status, test error information, and test performance indicators, the test results and the next state after the action is executed are obtained, and the adaptive reward process is obtained based on the test results and state changes.

[0081] After receiving the action instructions, the test execution and monitoring platform monitors the status, errors and performance of the high-speed interface;

[0082] Compare the interface test status obtained by monitoring with the next status to obtain the status distance reward;

[0083] Evaluate the severity and novelty of the monitored test error information and obtain error rewards;

[0084] Compare each test performance indicator obtained through monitoring with the performance benchmark value and obtain performance rewards;

[0085] The adaptive reward is obtained by weighted summing the state distance reward, error reward and performance reward;

[0086] Furthermore, the adaptive reward can be expressed as:

[0087] ;

[0088] in, is the time step Adaptive rewards under It is the status distance reward; Reward for mistakes; Reward for performance;

[0089] Furthermore, the state distance reward can be expressed as:

[0090] ;

[0091] in, It is the status distance reward; is the state reward weight; is the number of parameter types in the state space; is the Euclidean norm; is the time step The current update status of the following; is the time step The next state under

[0092] Furthermore, the error reward can be expressed as:

[0093] ;

[0094] ;

[0095] in, Reward for mistakes; Assess weights for severity; is the severity score; Assessing weight for novelty; is the novelty score;

[0096] Furthermore, the severity assessment weight and novelty assessment weight are set to 0.5 respectively; the weight setting can be adjusted according to actual needs;

[0097] Furthermore, the severity score is a normalized value assigned based on the error type, for example, 0.8 for link loss, 0.6 for a large number of CRC errors, and 0.4 for a small number of correctable errors. The novelty score is based on whether the state-action combination that caused the error is discovered for the first time, and is set to 1 if so, otherwise 0.

[0098] Furthermore, the performance reward can be expressed as:

[0099] ;

[0100] in, Reward for performance; reward weights for performance; is the number of performance parameters; For the The expected value of each performance parameter; For the The measured value of a performance parameter.

[0101] By comprehensively considering the state, errors, and performance indicators of high-speed interfaces, this adaptive reward can evaluate the value of state exploration, the severity and novelty of errors, and the actual deviation of performance. This helps the auxiliary agent learn better testing strategies, thereby ensuring the effectiveness and reliability of the high-speed interface fault-tolerant testing process.

[0102] Furthermore, the experience tuple includes: updated state, action, adaptive reward and next state.

[0103] In this embodiment, the experience tuple contains rich state prediction information and refined reward signals, which enables the reinforcement learning agent to be fully applied to complex fault-tolerant testing scenarios, making the adaptive testing process learn faster, explore more intelligently, and produce more effective results, thereby ensuring the effectiveness and reliability of the high-speed interface fault-tolerant testing process.

[0104] This embodiment provides a method for adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning. First, high-speed interface data, environmental data, and historical fault-tolerant test data are obtained; the fault-tolerant test impact prediction module is trained using the historical fault-tolerant test data to obtain a pre-training module and historical fault-tolerant test impact weights; based on the high-speed interface data and environmental data, the state, action, strategy, and reward elements of the reinforcement learning agent are defined; and a high-speed interface adaptive test is performed by combining the reinforcement learning agent, the pre-training module, and the test execution and monitoring platform. During the test process, the pre-training module is used to update the historical fault-tolerant test impact weights and the current state; based on the test results and state changes obtained by the test execution and monitoring platform through monitoring the interface test state, error information, and performance indicators, an adaptive reward is obtained; the present invention improves the effectiveness and reliability of the high-speed interface fault-tolerant test process.

[0105] Example 2:

[0106] This paper proposes a reinforcement learning-based method for adaptively generating high-speed interface fault-tolerant test processes. To further verify the effectiveness of the proposed fault-tolerant test impact prediction module and high-speed interface adaptive test process, the paper conducts module and process effectiveness tests for different module selections and test processes. Two companies, A and B, were selected for these two sets of tests.

[0107] The present invention selects the historical fault-tolerant test data of Company A over the past three years as the data set of the module, of which the data from the first two years is used as the training set of the model, and the data from the third year is used as the validation set. The sampling of the data set refers to the following rules: the data comes from the same interface configuration environment, and the data of four days of each week are randomly selected on a weekly basis; among them, three groups of data are extracted each day in the morning and afternoon.

[0108] The present invention inputs the training sets collected from Company A into different fault-tolerant test impact prediction modules for training to obtain respective pre-training modules; then inputs the validation sets into each pre-training module to obtain the impact prediction vector of each module; then, the impact prediction vector is compared with the actual situation through manual verification to obtain the proportion of the impact prediction vector of each module within a reasonable range.

[0109] The modules are: the fault-tolerant test impact prediction module proposed in the present invention, recorded as module one; the cross-modal interaction attention layer in the fault-tolerant test impact prediction module is removed, recorded as model two; the pooling operation in the feature fusion and compression layer is removed, and only the encoded features and the weighted features are fused, recorded as module three.

[0110] The module effectiveness test results are shown in Table 1.

[0111] Table 1 Module effectiveness test results

[0112] Test Module The proportion of the impact prediction vector within a reasonable range Model 1 90.59% Model 2 86.82% Model 3 88.78%

[0113] From the results in Table 1, it can be seen that the fault-tolerant test impact prediction module proposed in the present invention performs better than other modules in terms of effectiveness test results; thus, it can be seen that the module proposed in the present invention, combined with cross-modal interactive attention and multi-parameter feature fusion, can output accurate impact prediction vectors, further improving the effectiveness and reliability of the high-speed interface fault-tolerant test process.

[0114] To further verify the effectiveness of the high-speed interface adaptive test process, this example collected nearly a year's worth of historical data from Company B, sampling according to the same data sampling rules. The sampled data sets were then processed according to processes one, two, and three, respectively, to obtain their respective test results. Process one is the high-speed interface adaptive test process proposed in this invention; process two removes the state update operation of the pre-trained fault-tolerant test that affects the prediction module; process three calculates rewards solely based on state changes, without using the platform to monitor the interface. Finally, manual verification was performed to verify the effectiveness of each test process. The process effectiveness test results are shown in Table 2.

[0115] Table 2 Process effectiveness test results

[0116] Testing Process The proportion of test results within a reasonable range Process 1 90.43% Process 2 88.57% Process Three 87.92%

[0117] As can be seen from the results in Table 2, the effectiveness test results obtained by adopting process 1, that is, the high-speed interface adaptive test process proposed in the present invention, are better than those obtained by adopting other processes. This illustrates the necessity of combining the impact prediction vector and platform test results for high-speed interface adaptive testing, which can improve the effectiveness and reliability of the high-speed interface fault-tolerant test process.

[0118] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for adaptively generating high-speed interface fault-tolerant test processes based on reinforcement learning, characterized in that: include: Acquire high-speed interface data, environmental data, and historical fault-tolerance test data; The fault-tolerant test impact prediction module is trained using historical fault-tolerant test data to obtain a pre-trained fault-tolerant test impact prediction module and historical fault-tolerant test impact weights; Define the state, action, strategy, and reward elements of the reinforcement learning agent based on high-speed interface data and environmental data; High-speed interface adaptive testing is performed by combining a reinforcement learning agent, a pre-trained fault-tolerant test impact prediction module, and a test execution and monitoring platform. The testing process is as follows: setting the initial interface configuration and environmental conditions, and selecting the initial action based on prior knowledge of the interface; obtaining the current state, and using the pre-trained fault-tolerant test impact prediction module to update the historical fault-tolerant test impact weights and the current state to obtain the updated state; selecting an action based on the current strategy and the updated state, and having the test execution and monitoring platform execute the action instructions; Based on the platform's monitoring of interface test status, test error information, and test performance indicators, it obtains test results and the next state after executing the action; Obtain adaptive rewards based on test results and state changes; use experience tuples to update agent strategies; The loop iterates by continuously interacting with the environment and learning until the test reaches the preset condition.

2. The method for adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning according to claim 1, characterized in that: The high-speed interface data includes interface configuration parameter data and interface operation status data; the environmental data includes environmental parameter data and interference data; the historical fault-tolerant test data includes historical test configuration data, historical test process data and historical test result data.

3. The method for adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning according to claim 1, characterized in that: The process of training the fault tolerance test impact prediction module using the historical fault tolerance test data to obtain the pre-trained fault tolerance test impact prediction module and the historical fault tolerance test impact weights includes: Constructing the fault-tolerance test impact prediction module; Selecting historical test process data and historical test result data that match the configuration parameters of the interface to be tested from the historical fault-tolerant test data; wherein the historical test process data includes historical interface operation status data and historical environment data; The historical test process data and the historical test result data are input into the fault-tolerant test impact prediction module for processing to obtain the pre-trained fault-tolerant test impact prediction module and the historical fault-tolerant test impact weight under the interface configuration to be tested.

4. The method for adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning according to claim 1, characterized in that: The process of updating the historical fault-tolerant test impact weight and the current state using the pre-trained fault-tolerant test impact prediction module to obtain the updated state includes: Get the current state; Wherein, the current state includes the high-speed interface data and the environmental data; Inputting the interface operation status data, environmental parameter data and interference data in the current state into the pre-trained fault-tolerant test impact prediction module for processing, and updating the historical fault-tolerant test impact weight to generate an impact prediction vector; The impact prediction vector is combined with the current state to obtain the updated state.

5. The method for adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning according to claim 1, characterized in that: The platform monitors the interface test status, test error information, and test performance indicators, obtains test results and the next state after executing the action, and obtains adaptive rewards based on the test results and state changes. The process is as follows: The test execution and monitoring platform monitors the status, errors and performance of the high-speed interface after receiving the action instruction; Comparing the interface test state obtained by monitoring with the next state to obtain a state distance reward; Evaluate the severity and novelty of the monitored test error information and obtain error rewards; Compare each test performance indicator obtained through monitoring with the performance benchmark value and obtain performance rewards; A weighted sum is performed on the state distance reward, the error reward, and the performance reward to obtain the adaptive reward.

6. The method for adaptively generating a high-speed interface fault-tolerant test process based on reinforcement learning according to claim 1, characterized in that: The experience tuple includes: the updated state, the action, the adaptive reward, and the next state.

Images