Reverse test simulation method and device of vehicle, vehicle and storage medium

Abstract

The invention relates to the technical field of vehicle simulation testing, in particular to a reverse test simulation method and device of a vehicle, the vehicle and a storage medium. The method comprises the following steps: judging whether a reverse test simulation demand exists or not; under the condition that the reverse test simulation requirement exists, a vehicle simulation model is built, and a plurality of test scenes and a plurality of negative test cases meeting preset extreme conditions are generated through a preset knowledge graph; and loading the plurality of negative test cases into the vehicle simulation model in sequence, and performing a reverse test by using preset simulation software based on a plurality of test scenes to obtain a simulation result. Therefore, the virtual vehicle reverse test is carried out based on the knowledge graph, and the problems of insufficient reverse test capability and low test scene generation efficiency in the prior art are solved, so that the comprehensiveness of test cases is ensured, and the test scene generation efficiency and coverage rate are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle simulation testing, and particularly relates to a reverse test simulation method and device for a vehicle, a vehicle and a storage medium. BACKGROUND

[0002] With the rapid development of the automobile industry, vehicle testing has become a key link to ensure the safety and reliability of vehicles.

[0003] In the related art, virtual vehicle testing technology is gradually emerging, which simulates the behavior of a vehicle under various working conditions through computer simulation technology.

[0004] However, this method mainly focuses on forward testing, that is, it mainly verifies the performance of the system under ideal conditions, and is insufficient in reverse testing, and the test scene generation efficiency is low, which needs to be solved urgently. SUMMARY

[0005] The present application provides a reverse test simulation method and device for a vehicle, a vehicle and a storage medium to solve the problems of insufficient reverse testing capability and low test scene generation efficiency in the prior art, thereby ensuring the comprehensiveness of test cases and improving the efficiency and coverage of test scene generation.

[0006] To achieve the above-mentioned purpose, the first aspect of the present application provides a reverse test simulation method for a vehicle, comprising the following steps: determining whether there is a reverse test simulation requirement; in the case where the reverse test simulation requirement exists, building a vehicle simulation model, and generating a plurality of test scenes and a plurality of negative test cases satisfying a preset extreme condition by using a preset knowledge graph; loading the plurality of negative test cases into the vehicle simulation model in sequence, performing reverse testing by using a preset simulation software based on the plurality of test scenes, and obtaining a simulation result.

[0007] Through the above technical means, the present application realizes reverse testing of a vehicle under different extreme conditions, effectively improves the reverse testing capability, and at the same time, generates test scenes and negative test cases by using a preset knowledge graph, which significantly improves the efficiency of test scene generation.

[0008] According to one embodiment of the present application, before generating the plurality of test scenes and the plurality of negative test cases satisfying the preset extreme condition by using the preset knowledge graph, further comprising: acquire historical electrical inspection data and historical EOL (End of Line) detection system data, and respectively pre-process the historical electrical inspection data and the historical EOL detection system data to obtain pre-processed historical electrical inspection data and pre-processed historical EOL detection system data; determine a logical relationship between electrical appliance functions and at least one failure mode according to the pre-processed historical electrical inspection data and the pre-processed historical EOL detection system data; determine at least one entity and a relationship type between entities based on the logical relationship between the electrical appliance functions and the at least one failure mode, and construct an entity set based on the at least one entity and a relationship set based on the relationship type between the entities; import the entity set and the relationship set into a preset graph database to obtain an initial knowledge graph, and perform model training on the initial knowledge graph based on a preset multi-level decision graph and a preset loss function to obtain the preset knowledge graph.

[0009] Through the above technical means, comprehensive reverse test simulation of the vehicle electrical system can be realized, and test efficiency and accuracy can be improved. The historical electrical inspection data and the historical EOL detection system data provide rich data support for the training model, so that the model can better learn the logical relationship between electrical appliance functions and failure modes. At the same time, by importing the entity set and the relationship set into a preset graph database to construct an initial knowledge graph, and performing model training using a multi-level decision graph and a loss function, a more accurate and reliable knowledge graph can be obtained, which provides strong support for subsequent test simulation.

[0010] According to an embodiment of the present application, the pre-processing of the historical electrical inspection data and the historical EOL detection system data respectively to obtain pre-processed historical electrical inspection data and pre-processed historical EOL detection system data comprises: cleaning the historical electrical inspection data and the historical EOL detection system data respectively to obtain cleaned historical electrical inspection data and cleaned historical EOL detection system data; performing word segmentation processing on the cleaned historical electrical inspection data and the cleaned historical EOL detection system data using a preset pre-training model to obtain the pre-processed historical electrical inspection data and the pre-processed historical EOL detection system data.

[0011] Through the above technical means, the cleaning data process can remove noise, repeated values and abnormal values, etc., so that the data is more pure and consistent; and the word segmentation processing can split the text data into meaningful word units, which helps the model better understand the meaning and context information of the text. Such preprocessing steps not only improve the quality of the data, but also provide more accurate and rich information for subsequent model training, thereby improving the accuracy and reliability of the reverse test simulation.

[0012] According to an embodiment of the present application, before the model training based on the preset multi-level decision graph and the preset loss function is performed on the initial knowledge graph to obtain the preset knowledge graph, the following steps are further included: Based on the preset business logic, the hierarchical structure of the preset multi-level decision graph is determined. On each level, a decision node is constructed, and each decision node is connected based on the relationship type between the entities to obtain the preset multi-level decision graph.

[0013] Through the above technical means, a clear and hierarchical decision graph can be constructed, which can accurately reflect various decision relationships and business logic in the vehicle reverse test simulation process, not only helping to understand the overall process of the test simulation, but also providing strong support for subsequent model training and simulation optimization.

[0014] According to an embodiment of the present application, after the simulation result is obtained, the following steps are further included: According to the simulation result, a reverse test evaluation is performed to obtain a reverse test evaluation result.

[0015] Through the above technical means, the performance and safety of the vehicle can be more comprehensively and objectively evaluated, providing a strong basis for the design and improvement of the vehicle. At the same time, the reverse test evaluation result can also be used as a reference for subsequent simulation optimization and model training, further improving the accuracy and reliability of the reverse test simulation.

[0016] According to the vehicle reverse test simulation method proposed in the embodiments of the present application, in the case where there is a need for reverse test simulation, a vehicle simulation model is built, and a preset knowledge graph is used to generate a plurality of test scenarios and a plurality of negative test cases that meet preset extreme conditions. Subsequently, the plurality of negative test cases can be loaded into the vehicle simulation model one by one, and based on the plurality of test scenarios, a preset simulation software is used for reverse testing to obtain a simulation result. In this way, by using knowledge graph-based virtual vehicle reverse testing, the problems of insufficient reverse testing capability and low test scenario generation efficiency in the prior art are solved, thereby ensuring the comprehensiveness of the test cases and improving the efficiency and coverage of the test scenario generation.

[0017] To achieve the above object, the second aspect of the present application proposes a vehicle reverse test simulation device, comprising: A judgment module is configured to determine whether there is a reverse test simulation requirement. A generation module is configured to build a vehicle simulation model and generate a plurality of test scenarios and a plurality of negative test cases that meet preset extreme conditions using a preset knowledge graph if the reverse test simulation requirement exists. A test module is configured to load the plurality of negative test cases into the vehicle simulation model one by one, perform reverse testing based on the plurality of test scenarios using a preset simulation software, and obtain simulation results.

[0018] According to an embodiment of the present application, before the plurality of test scenarios and the plurality of negative test cases that meet the preset extreme conditions are generated using the preset knowledge graph, the generation module comprises: A preprocessing unit is configured to obtain historical electrical inspection data and historical EOL detection system data, and preprocess the historical electrical inspection data and the historical EOL detection system data respectively to obtain preprocessed historical electrical inspection data and preprocessed historical EOL detection system data. A determination unit is configured to determine a logical relationship between electrical appliance functions and at least one fault mode based on the preprocessed historical electrical inspection data and the preprocessed historical EOL detection system data. A construction unit is configured to determine at least one entity and a relationship type between entities based on the logical relationship between the electrical appliance functions and the at least one fault mode, construct an entity set based on the at least one entity, and construct a relationship set based on the relationship type between the entities. An obtaining unit is configured to import the entity set and the relationship set into a preset graph database to obtain an initial knowledge graph, and perform model training using the initial knowledge graph based on a preset multi-level decision graph and a preset loss function to obtain the preset knowledge graph.

[0019] According to an embodiment of the present application, the preprocessing unit is specifically configured to: The historical electrical inspection data and the historical EOL detection system data are cleaned respectively to obtain cleaned historical electrical inspection data and cleaned historical EOL detection system data. The cleaned historical electrical inspection data and the cleaned historical EOL detection system data are subjected to word segmentation processing using a preset pre-training model to obtain the preprocessed historical electrical inspection data and the preprocessed historical EOL detection system data.

[0020] According to one embodiment of the present application, before the model training based on the preset multi-level decision graph and the preset loss function is performed on the initial knowledge graph to obtain the preset knowledge graph, the obtaining unit is further configured to: determine a hierarchical structure of the preset multi-level decision graph based on a preset business logic; construct a decision node at each level, and connect each decision node based on the relationship type between the entities to obtain the preset multi-level decision graph.

[0021] According to one embodiment of the present application, after the simulation result is obtained, the testing module is further configured to: perform reverse testing evaluation based on the simulation result to obtain a reverse testing evaluation result.

[0022] According to the vehicle reverse testing simulation device provided by the embodiments of the present application, when there is a reverse testing simulation demand, a vehicle simulation model is built, and a plurality of test scenarios and a plurality of negative test cases satisfying a preset extreme condition are generated by using a preset knowledge graph. Then, the plurality of negative test cases can be loaded into the vehicle simulation model in sequence, and reverse testing is performed based on the plurality of test scenarios by using a preset simulation software to obtain a simulation result. Thus, by performing virtual whole vehicle reverse testing based on the knowledge graph, the problem of insufficient reverse testing capability and low test scenario generation efficiency in the prior art is solved, thereby ensuring the comprehensiveness of the test cases and improving the efficiency and coverage of the test scenario generation.

[0023] To achieve the above object, the third aspect of the present application provides a vehicle, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the reverse testing simulation method of the vehicle as described in the above embodiments.

[0024] To achieve the above object, the fourth aspect of the present application provides a computer readable storage medium having a computer program stored thereon, wherein the program is executed by a processor to implement the reverse testing simulation method of the vehicle as described in the above embodiments.

[0025] Additional aspects and advantages of the present application will be in part apparent and in part pointed out hereinafter. BRIEF DESCRIPTION OF DRAWINGS

[0026] The above and / or additional aspects and advantages of the present application will become apparent and be readily appreciated from the following description, taken in conjunction with the accompanying drawings, in which: Figure 1 A flowchart of a reverse testing simulation method of a vehicle according to an embodiment of the present application is provided. Figure 2 A block schematic diagram of a vehicle reverse test simulation device according to an embodiment of the present application is provided. Figure 3 A structural schematic diagram of a vehicle according to an embodiment of the present application is provided.

[0027] Label explanation: 10-vehicle reverse test simulation device, 100-judgment module, 200-generation module, 300-test module; 301-memory, 302-processor, 303-communication interface. DETAILED DESCRIPTION

[0028] The embodiments of the present application will be described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar notations represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present application, and cannot be understood as a limitation of the present application.

[0029] The vehicle reverse test simulation method, device, vehicle and storage medium according to the embodiments of the present application will be described below with reference to the accompanying drawings. First, the vehicle reverse test simulation according to the embodiments of the present application will be described with reference to the accompanying drawings.

[0030] Figure 1 A flowchart of a vehicle reverse test simulation method according to an embodiment of the present application.

[0031] Exemplarily, as shown in Figure 1 The vehicle reverse test simulation method includes the following steps: In step S101, it is judged whether there is a reverse test simulation demand.

[0032] It can be understood that reverse test refers to a test method for verifying the robustness and error handling capability of vehicle electrical system under extreme or abnormal conditions, and the core goal is to test the fault tolerance capability of the electrical system by simulating non-conventional working conditions (such as hardware failure, software anomaly or environmental mutation). Unlike traditional forward test (verifying the performance of electrical system under ideal conditions), reverse test actively injects abnormal parameters, for example, simulating sensor data out-of-range (such as -40℃ low temperature or 150℃ high temperature signal); manufacturing bus communication delay / interruption scene, etc.

[0033] In step S102, in the case where there is a reverse test simulation demand, a vehicle simulation model is built, and a plurality of test scenes and a plurality of negative test cases satisfying preset extreme conditions are generated by using a preset knowledge graph.

[0034] The knowledge graph is a structured data form for representing knowledge, usually composed of nodes and edges, which can be used to store complex logical relationships. In the embodiments of the present application, the knowledge graph is used to generate diversified test scenarios. Extreme conditions refer to environmental or operating conditions that exceed the normal range, such as extreme cold, high temperature, high load, etc., which are often used to test the performance of the system under extreme conditions. Negative test cases refer to test cases specifically designed to verify whether the system can correctly handle error inputs or abnormal situations, with the purpose of discovering vulnerabilities or defects of the system.

[0035] Specifically, in the case of reverse test simulation requirements, in order to meet this requirement, a vehicle simulation model can be first built. This model not only needs to accurately reflect the physical characteristics and behavior of the real vehicle, but also must be able to adapt to various complex test environments, thereby helping testers to discover potential problems and defects. In order to further enhance the comprehensiveness and depth of the test, the embodiments of the present application can use the preset knowledge graph to generate a series of test scenarios that meet certain extreme conditions (i.e. reverse test scenarios). These scenarios are designed to simulate edge cases that may be overlooked in regular testing, thereby ensuring that the vehicle system can maintain its performance and reliability when facing various extreme conditions. In addition, a plurality of negative test cases can be generated, which are specifically used to test the limit state of the electrical system, to ensure that the system still works as expected in the most unfavorable conditions.

[0036] Based on the information in the preset knowledge graph, a series of instructions can be designed, which can cover the response of the vehicle electrical system under various possible abnormal conditions. These instructions not only include normal operation instructions, but also include instructions that intentionally introduce errors or abnormalities, ensuring the comprehensiveness of the test cases to test the robustness and error handling capability of the electrical system.

[0037] By using the knowledge graph, automatically generating reverse test scenarios and negative test cases can cover more extreme working conditions, significantly improve the efficiency of test scenario generation, and improve the comprehensiveness of the test. In addition, through virtual simulation technology, the dependence on physical testing can be reduced, significantly reducing the testing cost and improving the testing efficiency.

[0038] In step S103, a plurality of negative test cases are loaded into the vehicle simulation model one by one, and based on a plurality of test scenarios, reverse testing is performed using a preset simulation software to obtain simulation results.

[0039] Specifically, based on multiple test scenarios and multiple negative test cases, the multiple negative test cases can be loaded into the vehicle simulation model in turn, and then the preset simulation software (such as CarSim, etc., which is not limited here) is used to run simulation tests according to the predetermined test process and conditions. This is a method of verifying whether the vehicle behavior conforms to the expectation from the expected result. During the test process, the key parameters and state information of the vehicle electrical system can be monitored and recorded in real time, such as sensor data, controller input and output signals, actuator action feedback, etc. Through this reverse test, a series of simulation results can be obtained, which can help identify potential problems and risks, and thus make necessary adjustments and optimizations to the vehicle design to ensure the safety and performance of the final product.

[0040] For ease of understanding, how to obtain the preset knowledge graph is described in detail below.

[0041] In some embodiments, before generating multiple test scenarios and multiple negative test cases that meet the preset extreme conditions using the preset knowledge graph, the method further includes: obtaining historical electrical inspection data and historical EOL detection system data, and pre-processing the historical electrical inspection data and the historical EOL detection system data respectively to obtain pre-processed historical electrical inspection data and pre-processed historical EOL detection system data; determining the logical relationship between electrical functions and at least one failure mode according to the pre-processed historical electrical inspection data and the pre-processed historical EOL detection system data; determining at least one entity and the relationship type between entities based on the logical relationship between electrical functions and at least one failure mode, and constructing an entity set based on at least one entity and constructing a relationship set based on the relationship type between entities; importing the entity set and the relationship set into a preset graph database to obtain an initial knowledge graph, and using the initial knowledge graph to train a model based on a preset multi-level decision graph and a preset loss function to obtain the preset knowledge graph.

[0042] It can be understood that the historical electrical inspection data can include the diagnosis requirements, instruction sets and implementation flow records in the historical electrical inspection process. The historical EOL detection system data can include the verification requirements of the EOL detection system, which can cover the action requirements, result requirements and other related standards that need to be performed during the detection process. The entity refers to the abstract generalization of the category of things that have independent existence meaning in the preset knowledge graph. In the embodiments of the present application, sensors, controllers, instructions, failure modes, etc. can all be entities. The relationship type refers to the association manner and the nature of the interaction between different entities, which reflects the logical relationship and the interaction manner between entities, such as dependency relationship, trigger relationship, association relationship, etc.

[0043] Specifically, in the process of constructing the preset knowledge graph, first, a custom dataset, i.e., historical electrical inspection data and historical EOL detection system data, can be loaded, each piece of data containing at least one piece of text, and in each piece of text, each Chinese character and each number corresponds to a specific label tag. Since the original data often has various formats and uneven quality, preprocessing operations can be performed to obtain preprocessed historical electrical inspection data and preprocessed historical EOL detection system data. Subsequently, with the aid of natural language processing technology, knowledge elements such as entities and relationship types can be extracted from the preprocessed historical electrical inspection data and the preprocessed historical EOL detection system data, i.e., the logical relationships (such as dependency relationships, trigger relationships, etc.) between electrical appliances (various sensors, controllers, actuators, etc.) functions are determined through data mining technology, and at least one failure mode (such as sensor failure, communication failure, etc.) is identified. Based on the logical relationships between electrical appliance functions and at least one failure mode, at least one entity can be determined, and the relationship types between entities can be determined. Based on these entities, an entity set can be constructed, and based on the relationship types between entities, a relationship set can be constructed, which will provide an important basis for subsequent analysis and decision-making.

[0044] Next, the entity set and the relationship set are imported into a preset graph database (such as Neo4j, OrientDB, etc.) to construct an initial knowledge graph. Based on the preset multi-level decision graph and the preset loss function, model training is performed using the initial knowledge graph, and the preset knowledge graph can be obtained. The overall training process can be set to 15 rounds, and the training process of each round is as follows: A batch of data is taken out from the dataloader (a tool for loading and processing data sets in deep learning); a batch of data is input into the model for forward calculation; the forward calculation result is transmitted to the preset loss function to calculate the loss, and the forward calculation result is transmitted to the evaluation method to calculate the evaluation index (such as precision, recall, F1 value, etc.); the calculated loss is transmitted back to update the gradient; clear the gradient and repeat the above steps.

[0045] It should be noted that the effect of the current model training can be evaluated every epoch (in deep learning, an epoch refers to the process of the entire training data set being completely traversed by the model once), and the model with the highest score is finally saved to obtain the preset knowledge graph.

[0046] It should be noted that in the embodiments of the present application, the construction and maintenance of the preset knowledge graph is a dynamic process, that is, by monitoring the vehicle electrical inspection system, EOL detection system and other data sources in real time, new data can be obtained in time, which can include new test results, new fault modes, new diagnosis requirements, etc. Based on the newly obtained data, the newly discovered knowledge can be fused with the current knowledge graph, that is, for the newly added entities and relationship types, new nodes and edges can be created in the preset graph database; for the existing entities and relationships, the attribute information thereof can be updated or the association relationship thereof can be adjusted to reflect the latest knowledge state. Through the update operation (such as insertion, update, deletion, etc.) of the graph database, the new knowledge content is written into the graph database, so as to ensure the timeliness and accuracy of the preset knowledge graph.

[0047] The following will be described in detail how to preprocess the historical electrical inspection data and the historical EOL detection system data.

[0048] As a possible implementation manner, in some embodiments, the historical electrical inspection data and the historical EOL detection system data are respectively preprocessed to obtain preprocessed historical electrical inspection data and preprocessed historical EOL detection system data, including: respectively cleaning the historical electrical inspection data and the historical EOL detection system data to obtain cleaned historical electrical inspection data and cleaned historical EOL detection system data; performing word segmentation processing on the cleaned historical electrical inspection data and the cleaned historical EOL detection system data by using a preset pre-training model to obtain the preprocessed historical electrical inspection data and the preprocessed historical EOL detection system data.

[0049] Specifically, in the process of preprocessing the historical electrical inspection data and the historical EOL detection system data, the historical electrical inspection data and the historical EOL detection system data can be cleaned first, including removing duplicate records, filling missing values, and standardizing units, to obtain cleaned historical electrical inspection data and cleaned historical EOL detection system data. Subsequently, the cleaned historical electrical inspection data and the cleaned historical EOL detection system data are subjected to word segmentation processing using a pre-set pre-training model (such as ERNIE (Enhanced Representation through kNowledge Integration), which processes Chinese data with a single Chinese character as the basic processing unit). PaddleNLP (Paddle Natural Language Processing) has built-in tokenizers for various pre-training models. A tokenizer is a tokenizer or token generator, which mainly converts the original input text into a form that the model can accept. Therefore, the tester can specify the name of the pre-training model (such as ERNIE) to be used, and the system will automatically load the corresponding tokenizer. After loading the tokenizer, it can be used to perform word segmentation processing on the input text, converting the text into a form that the model can accept. For example, for a Chinese text, the tokenizer will segment it into single Chinese characters, convert each Chinese character into a corresponding token ID, and then combine these token IDs into a sequence as input to the model.

[0050] The following explains how to obtain the pre-set multi-level decision graph.

[0051] In some embodiments, before obtaining the pre-set knowledge graph by training the initial knowledge graph based on the pre-set multi-level decision graph and the pre-set loss function, it further includes: determining the hierarchical structure of the pre-set multi-level decision graph based on the pre-set business logic; constructing decision nodes on each level, and connecting each decision node based on the relationship type between entities to obtain the pre-set multi-level decision graph.

[0052] It can be understood that the preset business logic refers to the business process, rules and objectives related to testing, which reflects the internal relationship and operation sequence between each link in the reverse testing process. The hierarchical structure of the preset multi-level decision graph can be divided into top layer (test objectives and overall process), middle layer (function modules and operation sequence) and bottom layer (specific operation and exception handling). The decision nodes constructed in the top layer can reflect the overall objectives and starting point of reverse testing, define the entry conditions and overall process framework of reverse testing. For example, the top layer node can be "start whole vehicle reverse testing", the condition is that the test environment is ready, and the operation is to initialize the vehicle state and enter the next step of testing. The decision nodes constructed in the middle layer can be divided according to the function modules of the vehicle system, each module corresponds to one or more decision nodes, and defines the operation sequence and condition judgment of the module in the testing process. For example, in the power system test module, there can be multiple decision nodes, respectively corresponding to different test instruction sending and response monitoring, such as "send engine start instruction", "monitor engine speed response", etc. The order and condition judgment between these nodes are determined according to the function logic of the power system and the test requirements. The decision nodes constructed in the bottom layer involve specific test operations and exception handling mechanisms. For example, after sending a certain test instruction, the bottom layer node will define how to monitor the system response, judge whether the response is normal, and take what measures in the case of exception, such as "if sensor A does not return data, trigger backup sensor B" or "if the system reports an error, record the error information and enter the fault diagnosis process".

[0053] Specifically, in constructing the preset multi-level decision graph, first, the hierarchical structure of the graph can be determined, which is designed according to business requirements and decision flow, ensuring that it can cover all relevant decision points. Once the hierarchical structure is determined, the decision nodes on each level can be constructed, which represent the decisions to be made at a particular level. In order to ensure the coherence and practicality of the preset multi-level decision graph, each decision node needs to be connected based on the relationship types between entities. The relationship types between entities can be causal relationship, time sequence relationship, logical relationship, etc., which determine the connection mode between decision nodes. Thus, a preset multi-level decision graph that conforms to business logic and has practical application value can be obtained.

[0054] That is, by deeply analyzing the preset business logic, each link and decision point in the reverse testing process can be clearly sorted out, so that the hierarchical structure of the preset multi-level decision graph can be reasonably defined, and the decision graph can accurately reflect the business process and rules of reverse testing.

[0055] For example, a preset multi-level decision graph (DG) can be defined as a four-tuple: DG=(E,C,A,R), where E (a set of factual entities): includes physical / logical entities involved in the test system, such as hardware (e.g., sensors, controllers) and software modules; C (a set of conditions): covers state conditions during the operation of the electrical system (e.g., voltage threshold, fault code triggering condition); A (a set of operations): contains a set of executable test actions (e.g., signal injection, state reset, and other control instructions); and R={E-C-A}: a set of rule triple elements that satisfy the condition triggering operation, for example, when a certain sensor value exceeds the threshold, a specific diagnostic instruction is triggered.

[0056] During the reverse test execution process, with the help of the preset multi-level decision graph (DG=(E,C,A,R)), the system state can be monitored in real time and decisions can be made. For example, when a certain condition (C) is met, the corresponding operation (A) is triggered according to the rule triple (R), thereby realizing the automatic control and real-time intervention of the reverse test process.

[0057] In addition, in some embodiments, after obtaining the simulation result, the method further includes: performing reverse test evaluation according to the simulation result to obtain a reverse test evaluation result.

[0058] That is, after obtaining the simulation result, the performance, reliability, and safety of the vehicle electrical system, and other aspects can be evaluated based on the simulation result, combined with a series of evaluation methods and indicators, the results of the comprehensive evaluation indicators are evaluated, and the reverse test evaluation result is obtained, and the corresponding optimization suggestions are given. For example, whether the electrical system can detect faults and take correct measures within a specified time, or whether the response of the electrical system under error input meets the design requirements, etc. The reverse test evaluation result can provide important decision support for the vehicle design and development team, help understand the advantages and disadvantages of the electrical system, and thus make reasonable adjustments and optimizations in the subsequent design and development process.

[0059] According to the vehicle reverse test simulation method provided in the embodiments of the present application, when there is a reverse test simulation demand, a vehicle simulation model is built, and a plurality of test scenarios and a plurality of negative test cases that meet preset extreme conditions are generated by using a preset knowledge graph. Subsequently, the plurality of negative test cases can be loaded into the vehicle simulation model in sequence, and based on the plurality of test scenarios, reverse testing is performed by using a preset simulation software to obtain a simulation result. Thus, by performing virtual whole vehicle reverse testing based on the knowledge graph, the problem of insufficient reverse testing capability and low test scenario generation efficiency in the prior art is solved, thereby ensuring the comprehensiveness of the test cases and improving the efficiency and coverage of the test scenario generation.

[0060] Next, a reverse test simulation device for a vehicle proposed in accordance with an embodiment of the present application will be described with reference to the accompanying drawings.

[0061] Figure 2 It is a block diagram of a reverse testing simulation device for a vehicle according to an embodiment of the present application.

[0062] like Figure 2 As shown, the vehicle reverse test simulation device 10 includes: a judgment module 100 , a generation module 200 and a test module 300 .

[0063] The judging module 100 is used to judge whether there is a reverse test simulation requirement; The generation module 200 is used to build a vehicle simulation model when there is a need for reverse test simulation, and use a preset knowledge graph to generate multiple test scenarios and multiple negative test cases that meet preset extreme conditions; The testing module 300 is used to sequentially load multiple negative test cases into the vehicle simulation model, and perform reverse testing based on multiple test scenarios using preset simulation software to obtain simulation results.

[0064] Optionally, in some embodiments, before using a preset knowledge graph to generate multiple test scenarios and multiple negative test cases that meet preset extreme conditions, the generation module 200 includes: a preprocessing unit, configured to obtain historical electrical inspection data and historical EOL detection system data, and preprocess the historical electrical inspection data and the historical EOL detection system data respectively to obtain preprocessed historical electrical inspection data and preprocessed historical EOL detection system data; a determination unit, configured to determine a logical relationship between electrical functions and at least one failure mode based on preprocessed historical electrical inspection data and preprocessed historical EOL detection system data; A construction unit, configured to determine, based on a logical relationship between electrical appliance functions and at least one failure mode, a type of relationship between at least one entity and the entities, and to construct an entity set based on the at least one entity, and to construct a relationship set based on the type of relationship between the entities; The acquisition unit is used to import the entity set and the relationship set into the preset graph database to obtain the initial knowledge graph, and based on the preset multi-level decision graph and the preset loss function, use the initial knowledge graph to perform model training to obtain the preset knowledge graph.

[0065] Optionally, in some embodiments, the preprocessing unit is specifically configured to: Cleaning the historical electrical inspection data and the historical EOL detection system data respectively to obtain cleaned historical electrical inspection data and cleaned historical EOL detection system data; The preprocessed historical electrical inspection data and the preprocessed historical EOL detection system data are obtained by using a preset pre-training model to perform word segmentation processing on the cleaned historical electrical inspection data and the cleaned historical EOL detection system data.

[0066] Optionally, in some embodiments, before the preset knowledge graph is obtained by using the initial knowledge graph to perform model training based on the preset multi-level decision graph and the preset loss function to obtain the preset knowledge graph, the obtaining unit is further configured to: determine a hierarchical structure of the preset multi-level decision graph based on a preset business logic; construct a decision node at each level, and connect each decision node based on a relationship type between entities to obtain the preset multi-level decision graph.

[0067] Optionally, in some embodiments, after the simulation result is obtained, the testing module 300 is further configured to: perform reverse testing evaluation according to the simulation result to obtain a reverse testing evaluation result.

[0068] It should be noted that the foregoing explanation and description of the embodiment of the reverse testing simulation method of the vehicle also apply to the reverse testing simulation device of the vehicle of this embodiment, which will not be described here again.

[0069] According to the reverse testing simulation device of the vehicle provided in the embodiments of the present application, when there is a demand for reverse testing simulation, a vehicle simulation model is built, and a preset knowledge graph is used to generate a plurality of test scenarios and a plurality of negative test cases that meet preset extreme conditions. Subsequently, the plurality of negative test cases can be loaded into the vehicle simulation model in sequence, and reverse testing is performed based on the plurality of test scenarios and using a preset simulation software to obtain a simulation result. In this way, virtual whole-vehicle reverse testing is performed based on a knowledge graph, which solves the problems of insufficient reverse testing capability and low efficiency of test scenario generation in the prior art, thereby ensuring the comprehensiveness of the test cases and improving the efficiency and coverage of test scenario generation.

[0070] Figure 3 A vehicle structure diagram is provided for the embodiments of the present application. The vehicle can include: a memory 301, a processor 302, and a computer program stored in the memory 301 and executable on the processor 302.

[0071] The processor 302 implements the reverse testing simulation method of the vehicle provided in the above embodiments when executing the program.

[0072] Further, the vehicle further includes: a communication interface 303 for communication between the memory 301 and the processor 302.

[0073] The memory 301 is configured to store a computer program executable in the processor 302.

[0074] The memory 301 can include a high-speed RAM (Random Access Memory) memory, and can further include a nonvolatile memory, for example, at least one disk memory.

[0075] If the memory 301, the processor 302 and the communication interface 303 are independently implemented, the communication interface 303, the memory 301 and the processor 302 can be connected to each other through a bus and complete communication between each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, Figure 3 Only one thick line is used in the figure, but it does not mean that there is only one bus or only one type of bus.

[0076] Optionally, in a specific implementation, if the memory 301, the processor 302 and the communication interface 303 are integrated on a chip, the memory 301, the processor 302 and the communication interface 303 can complete communication between each other through an internal interface.

[0077] The processor 302 can be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement one or more embodiments of the present application.

[0078] The embodiments of the present application also provide a computer readable storage medium, having stored thereon a computer program, which, when executed by a processor, implements the reverse test simulation method of the vehicle as above.

[0079] In addition, the terms "first", "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is at least two, for example, two, three, etc., unless otherwise explicitly and specifically limited.

[0080] In the description of the specification, the description using the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the particular feature, structure, material or characteristic being described is included in at least one embodiment or example of the present application. The illustrative appearances of the above-mentioned terms in various places in the specification are not necessarily referred to the same embodiment or example. Moreover, the particular features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples. Furthermore, the description herein of certain embodiments or examples does not necessarily exclude these embodiments or examples from the scope of the present application.

[0081] Although the embodiments of the present application have been shown and described above, it is understood that the above-mentioned embodiments are exemplary, and should not be understood as limiting the present application, and those ordinarily skilled in the art can make changes, modifications, replacements and variations to the above-mentioned embodiments within the scope of the present application.

Claims

1. A vehicle reverse test simulation method, characterized in that: The following steps are involved: Determine whether there is a need for reverse testing simulation; In the case of the above reverse test simulation requirements, a vehicle simulation model is built, and a preset knowledge graph is used to generate multiple test scenarios and multiple negative test cases that meet preset extreme conditions; The multiple negative test cases are sequentially loaded into the vehicle simulation model, and based on the multiple test scenarios, reverse testing is performed using preset simulation software to obtain simulation results.

2. The method according to claim 1, characterized in that Before using the preset knowledge graph to generate multiple test scenarios and the multiple negative test cases that meet the preset extreme conditions, the method further includes: Acquire historical electrical inspection data and historical EOL detection system data, and preprocess the historical electrical inspection data and the historical EOL detection system data respectively to obtain preprocessed historical electrical inspection data and preprocessed historical EOL detection system data; determining a logical relationship between electrical appliance functions and at least one failure mode based on the preprocessed historical electrical inspection data and the preprocessed historical EOL detection system data; Determining, based on the logical relationship between the electrical appliance functions and the at least one fault mode, a relationship type between at least one entity and entities, and constructing an entity set based on the at least one entity, and constructing a relationship set based on the relationship type between the entities; The entity set and the relationship set are imported into a preset graph database to obtain an initial knowledge graph, and based on a preset multi-level decision graph and a preset loss function, the initial knowledge graph is used to perform model training to obtain the preset knowledge graph.

3. The method according to claim 2, characterized in that The preprocessing of the historical electrical inspection data and the historical EOL detection system data to obtain the preprocessed historical electrical inspection data and the preprocessed historical EOL detection system data respectively includes: Cleaning the historical electrical inspection data and the historical EOL detection system data respectively to obtain cleaned historical electrical inspection data and cleaned historical EOL detection system data; The cleaned historical electrical inspection data and the cleaned historical EOL detection system data are word segmented using a preset pre-trained model to obtain the pre-processed historical electrical inspection data and the pre-processed historical EOL detection system data.

4. The method according to claim 2, characterized in that Before performing model training using the initial knowledge graph based on the preset multi-level decision graph and the preset loss function to obtain the preset knowledge graph, the method further includes: Based on the preset business logic, determine the hierarchical structure of the preset multi-level decision map; A decision node is constructed at each level, and based on the relationship type between the entities, each decision node is connected to obtain the preset multi-level decision graph.

5. The method according to claim 1, wherein After obtaining the simulation results, the method further includes: Perform a reverse test evaluation based on the simulation results to obtain a reverse test evaluation result.

6. A vehicle reverse test simulation device, characterized in that: include: A judgment module is used to judge whether there is a need for reverse test simulation; A generation module is used to build a vehicle simulation model when the reverse test simulation requirement exists, and use a preset knowledge graph to generate multiple test scenarios and multiple negative test cases that meet preset extreme conditions; The testing module is used to sequentially load the multiple negative test cases into the vehicle simulation model, and perform reverse testing based on the multiple test scenarios using preset simulation software to obtain simulation results.

7. The device according to claim 6, characterized in that Before using the preset knowledge graph to generate multiple test scenarios and the multiple negative test cases that meet the preset extreme conditions, the generation module includes: a preprocessing unit, configured to obtain historical electrical inspection data and historical EOL detection system data, and preprocess the historical electrical inspection data and the historical EOL detection system data respectively to obtain preprocessed historical electrical inspection data and preprocessed historical EOL detection system data; a determination unit, configured to determine a logical relationship between electrical appliance functions and at least one failure mode based on the preprocessed historical electrical inspection data and the preprocessed historical EOL detection system data; a construction unit, configured to determine, based on the logical relationship between the electrical appliance functions and the at least one fault mode, at least one entity and a type of relationship between entities, and to construct an entity set based on the at least one entity, and to construct a relationship set based on the type of relationship between the entities; An acquisition unit is used to import the entity set and the relationship set into a preset graph database to obtain an initial knowledge graph, and based on a preset multi-level decision graph and a preset loss function, use the initial knowledge graph to perform model training to obtain the preset knowledge graph.

8. The device according to claim 7, characterized in that The pre-processing unit is specifically used for: Cleaning the historical electrical inspection data and the historical EOL detection system data respectively to obtain cleaned historical electrical inspection data and cleaned historical EOL detection system data; The cleaned historical electrical inspection data and the cleaned historical EOL detection system data are word segmented using a preset pre-trained model to obtain the pre-processed historical electrical inspection data and the pre-processed historical EOL detection system data.

9. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the vehicle reverse test simulation method according to any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the vehicle reverse test simulation method according to any one of claims 1 to 5.

Images