New energy automobile controller test method, device and equipment based on natural language interaction and storage medium

Abstract

The invention discloses a new energy automobile controller test method, device and equipment based on natural language interaction and a storage medium, and relates to the technical field of automobile testing, and the method comprises the steps: taking a test natural language as the input of a vehicle test domain model, and obtaining a test target and test parameters; obtaining a structured scene description file based on the simulation interface knowledge graph, the test target and the test parameters; and generating and executing a test script based on the structured scene description file to obtain a test result. The technical problem of automatically and accurately converting a test intention described by a natural language into an executable test scene configuration signal in a simulation environment is solved. Compared with a mode depending on manual conversion in the prior art, full-automatic generation from natural language input to simulation scene configuration is achieved, a semantic gap between the test intention and simulation implementation is effectively eliminated, and the efficiency and accuracy of test scene construction are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of automotive testing technology, and in particular to a testing method, apparatus, equipment, and storage medium for new energy vehicle controllers based on natural language interaction. Background Technology

[0002] As the complexity of electronic control systems in new energy vehicles increases, model-based design processes require model-in-the-loop testing of the global controller to construct massive and complex test scenarios, thereby verifying the controller's performance under various extreme and coupled operating conditions. Therefore, there is a technical need to efficiently and accurately convert the test intentions described by testers in natural language into executable test scenario configuration signals in a simulation environment.

[0003] Currently, mainstream test scenario construction methods require test engineers to manually translate test intentions described in natural language into signal source module configurations, parameter settings, and bus signal mappings within specific simulation tools. This process heavily relies on engineers' deep understanding of model interfaces and simulation tools, resulting in technical drawbacks such as low scenario construction efficiency and a semantic gap between test intentions and simulation implementations.

[0004] Therefore, this technology needs to solve the technical problem of how to automatically and accurately convert test intentions described in natural language into executable test scenario configuration signals in a simulation environment, so as to achieve efficient and accurate test scenario construction.

[0005] The above content is only used to help understand the technical solution of this application and does not represent an admission that the above content is prior art. Summary of the Invention

[0006] The main purpose of this application is to provide a testing method, device, equipment and storage medium for new energy vehicle controllers based on natural language interaction, aiming to solve the technical problem of how to improve the efficiency of building vehicle testing scenarios.

[0007] To achieve the above objectives, this application proposes a testing method for new energy vehicle controllers based on natural language interaction, the method comprising: Acquire vehicle testing domain models, simulation interface knowledge graphs, and user-inputted test natural language; The test natural language is used as input to the vehicle test domain model to obtain the test target and test parameters; Based on the simulation interface knowledge graph, the test objective, and the test parameters, a structured scenario description file is obtained; Based on the structured scenario description file, a test script is generated; Execute the test script to obtain the test results.

[0008] In one embodiment, obtaining the simulation interface knowledge graph includes: Obtain the vehicle controller model; Based on the input and output interfaces of the vehicle controller model, determine the controller interface signals; Based on the controller interface signals, the signal name, data type, physical unit, and bus in which it is located are obtained; Establish the correspondence between the controller interface signals, the signal names, the data types, the physical units, and the buses they belong to, and obtain the simulation interface knowledge graph.

[0009] In one embodiment, obtaining a structured scenario description file based on the simulation interface knowledge graph, the test objective, and the test parameters includes: Based on the simulation interface knowledge graph and the test parameters, the model of the controller under test and the test signal are determined. Obtain the initial state corresponding to the test signal; Based on the test objective and the controller under test model, determine the timing events and key parameters; Based on the initial state, the timing events, and the key parameters, a structured scene description file is obtained.

[0010] In one embodiment, generating a test script based on the structured scenario description file includes: Based on the initial state in the structured scenario description file, the simulation parameters of the controller model under test are determined. Based on the timing events in the structured scene description file, the connection order of the signal source module under test and the signal source module under test is determined. Based on the key parameters in the structured scenario description file, a data recording scheme is determined; A test script is generated based on the simulation parameters of the controller model under test, the signal source module under test, the connection order of the signal source module under test, and the data recording scheme.

[0011] In one embodiment, executing the test script to obtain test results includes: Based on the simulation parameters of the controller model under test according to the test script, configure the initial parameters of the controller model under test. Based on the test script and the connection order of the signal source module under test, create and connect the signal source module under test; Based on the connected signal source module under test, a timing signal is generated; The timing signals of the test script are used as inputs to the controller under test model for model simulation. During the simulation, key signals are captured according to the data recording scheme of the test script to obtain the test results.

[0012] In one embodiment, obtaining a vehicle testing domain model includes: Obtain training instructions and simulation script examples in the field of new energy vehicles; Based on the training instructions for the new energy vehicle field and the simulation script examples, a large language model is trained to obtain a vehicle testing domain model.

[0013] In one embodiment, after executing the test script and obtaining the test results, the method further includes: The test results were quantitatively analyzed to obtain the quantitative analysis results; The quantitative analysis results are input into the vehicle testing domain model to obtain a natural language test report.

[0014] Furthermore, to achieve the above objectives, this application also proposes a new energy vehicle controller testing device based on natural language interaction, the new energy vehicle controller testing device based on natural language interaction comprising: The input module is used to acquire vehicle testing domain models, simulation interface knowledge graphs, and user-inputted test natural language. The recognition module is used to take the test natural language as input to the vehicle test domain model to obtain the test target and test parameters; The scenario module is used to obtain a structured scenario description file based on the simulation interface knowledge graph, the test target, and the test parameters; The script module is used to generate test scripts based on the structured scenario description file; The testing module is used to execute the test script and obtain the test results.

[0015] Furthermore, to achieve the above objectives, this application also proposes a new energy vehicle controller testing device based on natural language interaction. The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the new energy vehicle controller testing method based on natural language interaction as described above.

[0016] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the new energy vehicle controller testing method based on natural language interaction as described above.

[0017] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the new energy vehicle controller testing method based on natural language interaction as described above.

[0018] This application acquires a vehicle testing domain model, a simulation interface knowledge graph, and user-inputted test natural language; uses the test natural language as input to the vehicle testing domain model to obtain test objectives and test parameters; based on the simulation interface knowledge graph, the test objectives, and the test parameters, a structured scenario description file is obtained; based on the structured scenario description file, a test script is generated; and the test script is executed to obtain test results. By employing a vehicle testing domain model obtained through domain fine-tuning of a large language model as the core of semantic parsing, and combining it with a structured simulation interface knowledge graph for precise mapping, this application solves the technical problem of automatically and accurately converting test intentions described in natural language into executable test scenario configuration signals in the simulation environment. Compared with existing technologies that rely on manual conversion, this application achieves fully automated generation from natural language input to simulation scenario configuration, effectively eliminating the semantic gap between test intentions and simulation implementation, and significantly improving the efficiency and accuracy of test scenario construction. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating an embodiment of the testing method for a new energy vehicle controller based on natural language interaction provided in this application. Figure 2 This is a schematic diagram of the module structure of a new energy vehicle controller testing device based on natural language interaction, as described in an embodiment of this application. Figure 3 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the testing method for a new energy vehicle controller based on natural language interaction in the embodiments of this application.

[0022] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0023] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0024] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0025] The main solution of this application embodiment is as follows: acquiring a vehicle testing domain model, a simulation interface knowledge graph, and user-inputted test natural language; using the test natural language as input to the vehicle testing domain model to obtain test objectives and test parameters; obtaining a structured scenario description file based on the simulation interface knowledge graph, the test objectives, and the test parameters; generating a test script based on the structured scenario description file; and executing the test script to obtain test results.

[0026] In this embodiment, for ease of description, the following description will use the computer as the execution subject.

[0027] Mainstream test scenario construction methods require test engineers to manually translate test intentions described in natural language into specific signal source module configurations, parameter settings, and bus signal mappings within simulation tools. This process heavily relies on engineers' deep understanding of model interfaces and simulation tools, resulting in technical drawbacks such as low scenario construction efficiency and a semantic gap between test intentions and simulation implementations.

[0028] This application provides a solution that, by employing a vehicle testing domain model obtained through domain fine-tuning of a large language model as the core of semantic parsing, and combining it with a structured simulation interface knowledge graph for precise mapping, solves the technical problem of automatically and accurately converting test intentions described in natural language into executable test scenario configuration signals in the simulation environment. Compared with existing technologies that rely on manual conversion, this achieves fully automated generation from natural language input to simulation scenario configuration, effectively eliminating the semantic gap between test intentions and simulation implementation, and significantly improving the efficiency and accuracy of test scenario construction.

[0029] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or computer capable of performing the above functions. The following description uses a computer as an example to illustrate this embodiment and the subsequent embodiments.

[0030] Based on this, embodiments of this application provide a testing method for new energy vehicle controllers based on natural language interaction, referring to... Figure 1 , Figure 1This is a flowchart illustrating the first embodiment of the new energy vehicle controller testing method based on natural language interaction according to this application.

[0031] In this embodiment, the new energy vehicle controller testing method based on natural language interaction includes steps S10~S50: Step S10: Obtain the vehicle testing domain model, simulation interface knowledge graph, and user-inputted test natural language; It should be noted that the vehicle testing domain model refers to a large language model that has been fine-tuned with professional knowledge in the field of new energy vehicles, and has the ability to understand the testing intent and extract relevant testing elements; the simulation interface knowledge graph refers to a structured knowledge representation that contains detailed information on various input / output interface signals of the controller model under test and their semantic mapping relationships; and the user-input test natural language refers to the unstructured test scenario requirement text described by testers using natural language.

[0032] Understandably, since the test scenario construction process relies on a precise understanding of the simulation model interface, and directly using unverified or unstructured natural language input can easily lead to semantic ambiguity or mapping errors, step S10 can avoid information discrepancies between intent understanding and simulation implementation, thereby improving the accuracy and reliability of subsequent scenario parsing and script generation.

[0033] In one feasible implementation, step S10, which involves obtaining the simulation interface knowledge graph, includes: obtaining a vehicle controller model; determining controller interface signals based on the input and output interfaces of the vehicle controller model; obtaining the signal name, data type, physical unit, and bus based on the controller interface signals; and establishing the correspondence between the controller interface signals, the signal name, the data type, the physical unit, and the bus to obtain the simulation interface knowledge graph.

[0034] It should be noted that the vehicle controller model refers to the mathematical or software model of each control unit in a new energy vehicle within a simulation environment; the input interface refers to the port through which the model receives external signals, and the output interface refers to the port through which the model sends signals to the outside; the controller interface signal refers to the signal transmitted through these interfaces, representing a specific physical quantity or logical state; the signal name refers to the unique identifier of the signal in the model; the data type refers to the numerical type of the signal (such as Boolean, integer, or floating-point); the physical unit refers to the unit of measurement of the physical quantity represented by the signal; and the bus refers to the bus type to which the signal belongs in the vehicle communication network. The correspondence between controller interface signals, signal names, data types, physical units, and the bus refers to the structured recording of the relationships between the above information in a knowledge graph.

[0035] For example, for a BMS (Battery Management System) model, its input interface may include a "Battery_Temperature" signal, with data type float and unit °C, located on the CAN bus; the output interface may include a "SOC_Estimation" signal, with data type float and unit %, also located on the CAN bus. This information will be structured and stored in a knowledge graph.

[0036] In this embodiment, by structuring the interface information of the controller model into a knowledge graph, a precise and unambiguous mapping basis for the test parameters extracted from natural language to specific signals in the simulation environment is provided, solving the problems of easy errors and low efficiency when manually searching and configuring signals.

[0037] In one feasible implementation, step S10, obtaining the vehicle testing domain model, includes: obtaining training instructions and simulation script examples for the new energy vehicle domain; training a large language model based on the training instructions and simulation script examples for the new energy vehicle domain to obtain the vehicle testing domain model.

[0038] It should be noted that training instructions in the field of new energy vehicles refer to a set of text instructions containing professional corpora such as Autosar (Automotive Open System Architecture) terminology, vehicle dynamics parameters, and fault codes; simulation script examples refer to automated scripts of typical test scenarios that have been successfully executed and exist in the form of MATLAB scripts or Simulink test sequences; and large language models refer to pre-trained language models with powerful natural language understanding and generation capabilities.

[0039] For example, a large amount of natural language descriptions containing keywords such as "rapid acceleration," "battery SOC," and "gradient" and their corresponding simulation script fragments are used as pairing data to supervise and fine-tune the large language model, enabling it to learn to associate colloquial descriptions such as "overtaking at full throttle" with specific throttle opening timing signal generation logic.

[0040] In this embodiment, the large language model is fine-tuned by using domain-specific corpora and script examples, enabling it to acquire professional knowledge in the field of new energy vehicle testing. This solves the problem of inaccurate intent recognition and parameter extraction errors caused by the lack of domain knowledge when the general large language model parses test requirements.

[0041] The above are merely feasible implementations of step S10 provided in this embodiment. This embodiment does not specifically limit the specific implementation of step S10.

[0042] Step S20: Use the test natural language as input to the vehicle test domain model to obtain the test target and test parameters; It should be noted that the test objective refers to the core function, performance, or logical scenario to be verified in this test, identified from the natural language description. It is usually represented as an abstract test intent or test category. Test parameters refer to the specific quantitative values ​​or condition settings extracted from the natural language description required to achieve the test objective. These parameters will be used in subsequent simulation configurations.

[0043] For example, for a user-input natural language description of "simulating overtaking at full throttle on a long uphill road at -10 degrees Celsius and with the battery SOC at 50%, while the air conditioning is set to maximum heating," after analysis by the vehicle testing domain model: The test objective can be identified as: powertrain and thermal management coupling test.

[0044] The test parameters can be extracted as follows: ambient temperature = -10°C, battery SOC = 50%, road slope > 0%, throttle opening = 100%, air conditioning mode = maximum heating.

[0045] Understandably, due to the ambiguity and unstructured nature of natural language descriptions, directly mapping them to simulation signals can easily lead to misunderstandings or missing information. Therefore, step S20, which uses a vehicle testing domain model finely tuned with domain knowledge to perform structured parsing of the input, can avoid misjudgment of intent and parameter omissions caused by human intervention or simple rule-based parsing. This improves the accuracy and completeness of subsequent scenario construction, ensuring that the simulation test accurately reflects the user's original testing intent.

[0046] Step S30: Based on the simulation interface knowledge graph, the test target, and the test parameters, a structured scene description file is obtained; It should be noted that a structured scenario description file is a file recorded in a machine-readable standardized format (such as JSON or XML). It fully defines the simulation configuration information of the test scenario, including but not limited to the initial conditions of the simulation, the sequence of key events to be triggered on the timeline, and the signal and parameter values ​​corresponding to each event.

[0047] Understandably, since the test targets and parameters extracted from natural language are abstract and unstructured semantic information, they cannot be directly understood and executed by the simulation engine. Therefore, step S30 is performed to accurately bind the semantic parameters with specific simulation model interfaces and signals by utilizing the simulation interface knowledge graph. This can avoid simulation failure or result distortion caused by signal matching errors or timing logic confusion during the simulation configuration process. In this way, the test intentions expressed in natural language are efficiently and reliably transformed into structured driving instructions that can be recognized by the simulation environment.

[0048] In one feasible implementation, step S30 may include: determining the controller under test model and test signal based on the simulation interface knowledge graph and the test parameters; obtaining the initial state corresponding to the test signal; determining the timing events and key parameters based on the test target and the controller under test model; and obtaining a structured scene description file based on the initial state, the timing events, and the key parameters.

[0049] It should be noted that the controller under test (DUT) model refers to the specific controller model that needs to be verified and monitored in this test scenario, such as the VCU (Vehicle Control Unit) model. Test signals refer to the specific signals (such as voltage, temperature, and switching signals) that need to be input to the DUT model through the simulation interface to drive it or set its operating environment. The initial state corresponding to the test signal refers to the initial value or state that all relevant test signals should have at the start of the simulation (e.g., the ambient temperature signal is initially -10°C, and the battery SOC is initially 50%). Timing events refer to a series of operations or state changes that need to be triggered in a specific order on the simulation timeline (e.g., stepping the accelerator pedal signal from 0% to 100% at the 10th second). Key parameters define the specific values ​​or conditions for each timing event (e.g., the target value of the step is 100%, and the duration of the step).

[0050] For example, for the test target "low temperature charging test" and the test parameters "ambient temperature: -5°C, charging current: 50A": The tested controller models are defined as: BMS model and charger model.

[0051] The test signals were determined by querying the knowledge graph: BMS_AmbientTemp (ambient temperature signal) and Charger_OutputCurrent (charging current signal).

[0052] The initial state is set as follows: BMS_AmbientTemp = -5°C, Charger_OutputCurrent = 0A.

[0053] The timing event is defined as a "start charging" event that is triggered 2 seconds after the simulation begins.

[0054] The key parameter is: This event sets the Charger_OutputCurrent signal to 50A.

[0055] The final generated structured scene description file will contain all of the above information.

[0056] In this embodiment, by decomposing and mapping the abstract test objectives and parameters into specific models, signals, initial states, and timing events, a precise, complete, and executable scenario blueprint is constructed, which solves the problems of easily omitting initialization conditions, disordered timing logic, or inaccurate signal mapping when manually designing simulation processes.

[0057] The above are merely feasible implementations of step S30 provided in this embodiment. This embodiment does not specifically limit the specific implementation of step S30.

[0058] Step S40: Generate a test script based on the structured scene description file; It should be noted that a test script is an automated program file containing a series of commands that can be recognized and executed by simulation tools (such as MATLAB). Its function is to transform a structured scenario description into instructions that the simulation environment can directly run, including model initialization, signal generation and injection, simulation operation control, and data acquisition.

[0059] Understandably, since the structured scenario description file only describes the "blueprint" of the test scenario, while the simulation tool requires specific instruction sequences and configuration operations to execute the simulation, step S40, by automatically generating an executable test script, can avoid the syntax errors, logical errors, and inconsistencies with the model interface that are prone to occur when manually writing scripts. This ensures a high degree of consistency between the simulation process and the test intent, and significantly improves the efficiency and reliability of the test execution process.

[0060] In one feasible implementation, step S40 may include: determining the simulation parameters of the controller under test model based on the initial state in the structured scenario description file; determining the connection order of the signal source module under test and the signal source module under test based on the timing events in the structured scenario description file; determining the data recording scheme based on the key parameters in the structured scenario description file; and generating a test script based on the simulation parameters of the controller under test model, the signal source module under test, the connection order of the signal source module under test, and the data recording scheme.

[0061] It should be noted that the simulation parameters of the controller under test (DUT) model refer to the state variables or environmental parameter values ​​that need to be set in the DUT model or its associated models at the start of the simulation. The signal source module under test (SUTP) refers to the module in the simulation model used to generate specific timing signals to drive the inputs of the DUT. The connection order of the SUTP modules refers to how these signal source modules are combined in chronological order and how their output signals are connected to the corresponding input ports of the DUT model. The data logging scheme refers to the configuration of which simulation output signals need to be recorded, at what sampling rate, and in what format the log files are for subsequent analysis.

[0062] For example, in a scenario of "rapid acceleration to overtake," the test script includes the following: Simulation parameters: The initial vehicle speed is set to 80 km / h, and the road surface adhesion coefficient is set to 0.85.

[0063] The signal source module under test and its connection sequence: A cruise control signal source module outputs a constant desired speed signal of 80 km / h and is connected to the driver model.

[0064] At the 5th second, a "step throttle" signal source module is activated, instantly boosting the throttle opening signal from 10% to 100% and connecting it to the throttle input port of the VCU.

[0065] Data recording scheme: Record motor torque, battery current, and vehicle speed signals at a sampling rate of 100 Hz and save them as .mat files.

[0066] The final generated test script will be a MATLAB .m file, which contains code for setting the above parameters, creating and connecting the signal source module, configuring the logs, and starting the simulation.

[0067] In this embodiment, by systematically translating the various elements (states, events, parameters) in the structured scene description into specific configurations and instructions of the simulation environment, an executable script that strictly corresponds to the scene description is automatically generated, solving the problems of tedious, error-prone, time-consuming, and difficult-to-ensure consistency with the original scene intent when manually writing scripts.

[0068] The above are merely feasible implementations of step S40 provided in this embodiment. This embodiment does not specifically limit the specific implementation of step S40.

[0069] Step S50: Execute the test script to obtain the test results.

[0070] It should be noted that test results refer to the set of output data generated by the simulation model after executing the simulation test script, reflecting the behavior and performance of the controller under test in a specific scenario. This data typically includes time-series signals, event logs, performance metric calculation results, and simulation state information.

[0071] Understandably, since the ultimate goal of simulation testing is to obtain the controller's response data under preset operating conditions for evaluation, manually executing test scripts and collecting data is not only inefficient but also prone to data omissions or recording errors due to operational negligence. Therefore, step S50, by automating the scheduling and execution of test scripts and strictly recording simulation data according to the preset scheme, can avoid operational inconsistencies and data recording errors introduced by manual intervention, thereby ensuring the repeatability of the test process and the integrity and reliability of the result data.

[0072] In one feasible implementation, step S50 may include: configuring the initial parameters of the controller under test model based on the simulation parameters of the controller under test model in the test script; creating and connecting the signal source module under test based on the connection order of the signal source module under test in the test script; generating timing signals based on the connected signal source module under test; using the timing signals of the test script as input to the controller under test model for model simulation; and capturing key signals according to the data recording scheme of the test script during the simulation to obtain test results.

[0073] It should be noted that the initial parameters of the controller under test (DUT) model refer to the specific values ​​directly assigned to the model's internal variables or operating points before simulation (e.g., setting the initial SOC value of the battery in the BMS model, or setting the initial temperature of the coolant in the thermal management model). Timing signals refer to signals that change according to a specific time pattern, used to simulate changes in input quantities in the real world (e.g., a gradient curve signal that changes over time, or a brake pedal signal that experiences a step change at a specific moment). Model simulation refers to the process by which simulation tools calculate and output the evolution of the model's state and output signals over time based on the model's mathematical relationships and input signals. Critical signals refer to signals that require special attention and recording under the guidance of the test objectives, and are usually directly related to evaluation indicators (e.g., the motor torque response signal used to evaluate response speed, or the battery's maximum temperature signal used to evaluate safety).

[0074] For example, in the "low temperature fast charging" test scenario: Configure initial parameters: Before simulation, set the initial battery temperature of the BMS model to -10°C.

[0075] Create and connect the signal source module: Insert a "charging current" signal source module into the model and connect its output port to the current command input port of the charger model.

[0076] Generate timing signal: This signal source module generates a current signal that jumps from 0A to 150A and holds based on the script.

[0077] Model simulation: The Simulink engine begins calculations, and the BMS and battery model respond to the charging current input.

[0078] Capture key signals to obtain results: According to the recording scheme, record signals such as battery voltage, temperature, and SOC change curves in real time to form a test result dataset.

[0079] In this embodiment, by automatically and accurately applying the configuration information in the test script to the simulation environment and driving the simulation execution and data acquisition, the problems of cumbersome steps, error-prone configuration, and incomplete data acquisition when manually operating simulation software are solved, and a high degree of automation and standardization of the test execution process is achieved.

[0080] In one feasible implementation, step S50 may include: performing quantitative analysis on the test results to obtain quantitative analysis results; and inputting the quantitative analysis results into the vehicle test domain model to obtain a natural language test report.

[0081] It should be noted that quantitative analysis results refer to numerical values ​​with clear physical or engineering significance obtained after processing the raw test result data (time series signals) (e.g., the rise time of motor torque from 10% to 90% is 120ms, the maximum temperature rise of the battery during the test is 8°C, and the exact moment a fault code is triggered). Natural language test reports refer to test conclusions, phenomenon descriptions, performance evaluations, and potential problem analyses presented in natural language paragraphs that are easily read and understood by humans and are automatically generated by the system.

[0082] For example, based on the test engineer's input of the natural language "Please test the situation where the vehicle suddenly loses GPS signal while cruising at high speed, and the right front tire pressure is too low," a natural language test report is generated: "Test complete. In the simulated scenario, the vehicle switches to inertial navigation within 3 seconds after GPS loss, but the low tire pressure alarm and navigation degradation information appear simultaneously, which may confuse the driver. It is recommended to optimize the alarm information priority strategy."

[0083] In this implementation, by automatically extracting indicators from the raw data and using a domain-knowledge-based model to "translate" the data into natural language conclusions with explanations and insights, the problem of highly specialized test results data, high interpretation threshold, time-consuming and laborious manual report writing, and potential omission of key insights is solved, greatly improving the efficiency of test result interpretation and the effectiveness of knowledge transfer.

[0084] The above are merely feasible implementations of step S50 provided in this embodiment. This embodiment does not specifically limit the specific implementation of step S50.

[0085] This embodiment provides a testing method for new energy vehicle controllers based on natural language interaction. It acquires a vehicle testing domain model, a simulation interface knowledge graph, and user-inputted test natural language. The test natural language is used as input to the vehicle testing domain model to obtain test objectives and test parameters. Based on the simulation interface knowledge graph, the test objectives, and the test parameters, a structured scenario description file is obtained. A test script is generated based on the structured scenario description file. The test script is executed to obtain test results. By employing a vehicle testing domain model obtained through domain fine-tuning of a large language model as the core of semantic parsing, and combining this with a structured simulation interface knowledge graph for precise mapping, the technical problem of automatically and accurately converting test intentions described in natural language into executable test scenario configuration signals in the simulation environment is solved. Compared with existing technologies that rely on manual conversion, this method achieves fully automated generation from natural language input to simulation scenario configuration, effectively eliminating the semantic gap between test intentions and simulation implementation, and significantly improving the efficiency and accuracy of test scenario construction.

[0086] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the testing method for new energy vehicle controllers based on natural language interaction in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0087] This application also provides a new energy vehicle controller testing device based on natural language interaction, please refer to... Figure 2 The new energy vehicle controller testing device based on natural language interaction includes: Input module 10 is used to acquire vehicle testing domain model, simulation interface knowledge graph and test natural language input by user; The recognition module 20 is used to take the test natural language as input to the vehicle test domain model to obtain the test target and test parameters; Scenario module 30 is used to obtain a structured scenario description file based on the simulation interface knowledge graph, the test target, and the test parameters; Script module 40 is used to generate test scripts based on the structured scene description file; Test module 50 is used to execute the test script and obtain test results.

[0088] The new energy vehicle controller testing device based on natural language interaction provided in this application, employing the new energy vehicle controller testing method based on natural language interaction in the above embodiments, can solve the technical problem of how to improve the efficiency of constructing vehicle test scenarios. Compared with the prior art, the beneficial effects of the new energy vehicle controller testing device based on natural language interaction provided in this application are the same as the beneficial effects of the new energy vehicle controller testing method based on natural language interaction provided in the above embodiments, and other technical features in the new energy vehicle controller testing device based on natural language interaction are the same as the features disclosed in the methods of the above embodiments, and will not be repeated here.

[0089] The input module 10 is also used to acquire training instructions and simulation script examples in the field of new energy vehicles; and to train a large language model based on the training instructions and simulation script examples in the field of new energy vehicles to obtain a vehicle testing domain model.

[0090] The input module 10 is further configured to acquire a vehicle controller model; determine controller interface signals based on the input and output interfaces of the vehicle controller model; obtain the signal name, data type, physical unit, and bus based on the controller interface signals; and establish the correspondence between the controller interface signals, the signal name, the data type, the physical unit, and the bus to obtain a simulation interface knowledge graph.

[0091] The scenario module 30 is further configured to determine the controller under test model and test signal based on the simulation interface knowledge graph and the test parameters; obtain the initial state corresponding to the test signal; determine the timing events and key parameters based on the test target and the controller under test model; and obtain a structured scenario description file based on the initial state, the timing events and the key parameters.

[0092] The script module 40 is further configured to: determine the simulation parameters of the controller under test model based on the initial state in the structured scene description file; determine the connection order of the signal source module under test and the signal source module under test based on the timing events in the structured scene description file; determine the data recording scheme based on the key parameters in the structured scene description file; and generate a test script based on the simulation parameters of the controller under test model, the signal source module under test, the connection order of the signal source module under test, and the data recording scheme.

[0093] The test module 50 is further configured to configure the initial parameters of the controller model under test based on the simulation parameters of the controller model under test in the test script; create and connect the signal source module under test based on the connection order of the signal source module under test in the test script; generate timing signals based on the connected signal source module under test; use the timing signals of the test script as input to the controller model under test for model simulation; and capture key signals according to the data recording scheme of the test script during the simulation process to obtain test results.

[0094] The testing module 50 is also used to perform quantitative analysis on the test results to obtain quantitative analysis results; and input the quantitative analysis results into the vehicle testing domain model to obtain a natural language test report.

[0095] This application provides a new energy vehicle controller testing device based on natural language interaction. The new energy vehicle controller testing device based on natural language interaction includes: at least one processor; and a memory communicatively connected to at least one processor; wherein the memory stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor to enable at least one processor to execute the new energy vehicle controller testing method based on natural language interaction in the above embodiment 1.

[0096] The following is for reference. Figure 3 This document illustrates a structural schematic diagram of a new energy vehicle controller testing device suitable for implementing embodiments of this application based on natural language interaction. The new energy vehicle controller testing device based on natural language interaction in this application embodiment may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 3 The new energy vehicle controller test equipment based on natural language interaction shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0097] like Figure 3As shown, the new energy vehicle controller testing equipment based on natural language interaction may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to the program stored in ROM (Read Only Memory) 1002 or the program loaded from storage device 1003 into random access memory (RRAM) 1004. RAM 1004 also stores various programs and data required for the operation of the new energy vehicle controller testing equipment based on natural language interaction. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the new energy vehicle controller test equipment based on natural language interaction to exchange data wirelessly or via wired communication with other devices. Although the figure shows a new energy vehicle controller test equipment based on natural language interaction with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0098] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0099] The new energy vehicle controller testing equipment based on natural language interaction provided in this application, employing the new energy vehicle controller testing method based on natural language interaction in the above embodiments, can solve the technical problem of how to improve the efficiency of constructing vehicle test scenarios. Compared with the prior art, the beneficial effects of the new energy vehicle controller testing equipment based on natural language interaction provided in this application are the same as the beneficial effects of the new energy vehicle controller testing method based on natural language interaction provided in the above embodiments, and other technical features in this new energy vehicle controller testing equipment based on natural language interaction are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0100] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0101] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0102] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, which are used to execute the new energy vehicle controller testing method based on natural language interaction in the above embodiments.

[0103] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0104] The aforementioned computer-readable storage medium may be included in a new energy vehicle controller testing device based on natural language interaction; or it may exist independently and not be assembled into a new energy vehicle controller testing device based on natural language interaction.

[0105] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by a new energy vehicle controller testing device based on natural language interaction, the new energy vehicle controller testing device based on natural language interaction performs the following actions: acquires a vehicle testing domain model, a simulation interface knowledge graph, and test natural language input by the user; uses the test natural language as input to the vehicle testing domain model to obtain test objectives and test parameters; obtains a structured scenario description file based on the simulation interface knowledge graph, the test objectives, and the test parameters; generates a test script based on the structured scenario description file; and executes the test script to obtain test results.

[0106] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0107] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0108] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0109] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described new energy vehicle controller testing method based on natural language interaction, thereby solving the technical problem of how to improve the efficiency of constructing vehicle testing scenarios. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the new energy vehicle controller testing method based on natural language interaction provided in the above embodiments, and will not be repeated here.

[0110] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described test method for a new energy vehicle controller based on natural language interaction.

[0111] The computer program product provided in this application can solve the technical problem of how to improve the efficiency of constructing automotive test scenarios. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the new energy vehicle controller testing method based on natural language interaction provided in the above embodiments, and will not be repeated here.

[0112] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A new energy vehicle controller test method based on natural language interaction, characterized in that, The method comprises: acquiring a vehicle test field model, a simulation interface knowledge graph and user input test natural language; inputting the test natural language into the vehicle test field model to obtain a test target and a test parameter; obtaining a structured scene description file based on the simulation interface knowledge graph, the test target and the test parameter; generating a test script based on the structured scene description file; executing the test script to obtain a test result.

2. The method of claim 1, wherein, The simulation interface knowledge graph comprises: acquiring a vehicle controller model; determining a controller interface signal according to an input interface and an output interface of the vehicle controller model; obtaining a signal name, a data type, a physical unit and a bus based on the controller interface signal; establishing a correspondence between the controller interface signal, the signal name, the data type, the physical unit and the bus to obtain a simulation interface knowledge graph.

3. The method of claim 1, wherein, The simulation interface knowledge graph, the test target and the test parameter comprise: determining a tested controller model and a test signal based on the simulation interface knowledge graph and the test parameter; obtaining an initial state corresponding to the test signal; determining a timing event and a key parameter based on the test target and the tested controller model; obtaining a structured scene description file based on the initial state, the timing event and the key parameter.

4. The method of claim 1, wherein, The structured scene description file comprises: determining a simulation parameter of the tested controller model based on the initial state in the structured scene description file; determining a tested signal source module and a connection order of the tested signal source module based on the timing event in the structured scene description file; determining a data recording scheme based on the key parameter in the structured scene description file; generating a test script based on the simulation parameter of the tested controller model, the tested signal source module, the connection order of the tested signal source module and the data recording scheme.

5. The method of claim 1, wherein, The test script comprises: configuring an initial parameter of the tested controller model based on the simulation parameter of the tested controller model in the test script; creating and connecting a tested signal source module based on the tested signal source module and the connection order of the tested signal source module in the test script; generating a timing signal based on the connected tested signal source module; performing model simulation by taking the timing signal of the test script as input of the tested controller model; capturing a key signal according to the data recording scheme of the test script during simulation to obtain a test result.

6. The method of claim 1, wherein, The vehicle test field model comprises: acquiring a new energy vehicle field training instruction and a simulation script example; training a large language model based on the new energy vehicle field training instruction and the simulation script example to obtain a vehicle test field model.

7. The method of claim 1, wherein, After executing the test script to obtain a test result, the method further comprises: performing quantitative analysis on the test result to obtain a quantitative analysis result; The quantification analysis result is input into the vehicle test field model to obtain a natural language test report.

8. A new energy vehicle controller testing device based on natural language interaction, characterized in that, The device comprises: An input module configured to acquire a vehicle test field model, a simulation interface knowledge graph, and a test natural language input by a user; An identification module configured to input the test natural language into the vehicle test field model to obtain a test target and a test parameter; A scenario module configured to obtain a structured scenario description file based on the simulation interface knowledge graph, the test target, and the test parameter; A script module configured to generate a test script based on the structured scenario description file; A test module configured to execute the test script to obtain a test result.

9. A new energy vehicle controller test device based on natural language interaction, characterized in that, The device comprises a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the natural language interaction-based new energy vehicle controller test method according to any one of claims 1 to 7.

10. A storage medium, characterized by The storage medium is a computer-readable storage medium, and the storage medium stores a computer program, and the computer program is executed by the processor to implement the steps of the natural language interaction-based new energy vehicle controller test method according to any one of claims 1 to 7.

Images