System and method for testing man-machine interaction function of intelligent driving assistance system of heavy-duty car

Abstract

The invention discloses a system and method for testing the man-machine interaction function of an intelligent driving assistance system of a heavy-duty car, and belongs to the technical field of intelligent testing of cars, and the system comprises a data collection module, an analysis processing module, a data comparison module and a data statistics module. The data acquisition module is in communication connection with the analysis processing module and is used for acquiring multi-source data in the man-machine interaction process of the heavy-duty car; the analysis processing module is in communication connection with the data comparison module and is used for receiving the multi-source data and processing the multi-source data; the data comparison module is in communication connection with the data statistics module and is used for comparing the processing result with a corresponding test case in a standardized test case library and judging whether the processing result meets a preset standard or not; and the data statistics module is used for receiving the comparison judgment result, outputting a test conclusion and an optimization suggestion and generating a statistical chart. The method is suitable for function verification and performance optimization test of the vehicle-mounted man-machine interaction system of the heavy-duty vehicle, and provides test support for safe landing of the intelligent driving system.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent testing of automobiles, and particularly relates to a heavy truck intelligent driving assistance system man-machine interaction function test system and method. BACKGROUND

[0002] With the intelligent upgrading of heavy trucks, man-machine interaction functions (such as voice control gear shifting, central control screen operation response, instrument panel information feedback, etc.) become the core component of the intelligent driving system, and the function stability thereof directly affects driving safety.

[0003] In the prior art, a test personnel manually records the response result of the man-machine interaction function through real vehicle operation, and relies on the experience of the test personnel, which is highly subjective. Different test personnel may obtain different results for the same function, vehicle-mounted message data cannot be synchronously collected, it is difficult to determine whether a function abnormality is caused by an interaction module or a vehicle-mounted ECU, test scene coverage is limited, concurrent scenes such as continuous voice instructions and complex road conditions cannot be simulated, and the like. Part of the interaction data is obtained through a single data collection device (such as a special camera for collecting images and a special microphone for collecting sound), and then manually imported into a computer for comparison. The data collection is not synchronous, the time stamp deviation of image, sound and message data can reach hundreds of milliseconds, and problems such as normal display but delayed message cannot be detected. There is a lack of unified data analysis and statistics module, the test result needs to be manually sorted and a report needs to be generated, which is low in efficiency. Test cases cannot be dynamically updated, and the flexibility is poor. SUMMARY

[0004] The application aims to solve the problems of easy omission of test results, low test efficiency and limited scene coverage in the prior art, and provides a heavy truck intelligent driving assistance system man-machine interaction function test system and method.

[0005] In a first aspect, the application provides a heavy truck intelligent driving assistance system man-machine interaction function test system, which comprises: a data collection module, an analysis processing module, a data comparison module and a data statistics module; The data collection module is in communication connection with the analysis processing module. The data collection module is configured to synchronously collect multi-source data in a heavy truck man-machine interaction process. The multi-source data comprises perception data and message data. The perception data comprises image data, sound data and vibration data. The message data is interaction instruction data transmitted by a heavy truck vehicle-mounted system. The analysis processing module is in communication connection with the data comparison module, the analysis processing module is used for receiving the multi-source data transmitted by the data acquisition module, and the multi-source data is processed, the processing operation includes feature extraction of image data, filtering processing and extraction of sound data, and format analysis and validity check of message data, and the processing result is output; The data comparison module is in communication connection with the data statistical module, the data comparison module pre-stores a standardized test case library, the data comparison module is used for comparing the processing result output by the analysis processing module with the corresponding test case in the standardized test case library one by one, judging whether the processing result meets the preset standard, and generating comparison judgment result; The data statistical module is used for receiving the comparison judgment result, outputting the test result including test pass or fail and optimization suggestion in real time, and automatically generating statistical charts including test pass rate and problem distribution after a single test process is completed.

[0006] Further, the data acquisition module includes an image acquisition unit, a sound acquisition unit, a message acquisition unit and a vibration acquisition unit; The image acquisition unit is a high-definition industrial camera, used for acquiring image data, and the image data is a central control screen image; The sound acquisition unit is a noise reduction microphone array, used for acquiring sound data, and the sound data is a system prompt tone or a voice instruction; The message acquisition unit includes a CAN bus interface, and the message acquisition unit is used for acquiring message data; The vibration acquisition unit is used for acquiring vibration data, and the vibration data includes vibration information of a seat, a safety belt and a steering wheel.

[0007] Further, the analysis processing module is built-in image recognition algorithm and sound recognition algorithm; The image recognition algorithm is used for extracting display element position, text information and state identifier in image data; The sound recognition algorithm is used for filtering environmental noise in sound and recognizing instruction content and prompt tone type.

[0008] Further, the standardized test case library in the data comparison module is added, modified and deleted according to the vehicle model upgrade and man-machine interaction function iteration demand of heavy trucks; Each test case includes four fields of scene number, test step, expected result and judgment threshold; The judgment threshold is in a standardized format, including instrument text, central control text, instrument icon, central control icon, sound, voice and vibration parameters.

[0009] Further, the statistical chart output by the data statistics module includes a column chart, a line chart, and a pie chart. The column chart is used to show the number of test passes and the number of test failures in different interaction scenarios. The line chart is used to show the trend of test results of the same scenario in multiple retests. The pie chart is used to show the distribution of the proportion of various types of test problems.

[0010] In a second aspect, the embodiments of the present application also provide a heavy truck intelligent driving assistance system human-computer interaction function test method, which comprises the following steps: S1. Build a real-time data acquisition link compatible with the heavy truck on-board system, so that the data acquisition module is stably connected to the heavy truck data transmission channel; according to the heavy truck human-computer interaction function test requirements, write standardized test cases covering key operation response, voice command recognition, display interface feedback, and vibration feedback of the interaction scenario, and store the standardized test cases in the standardized test case library of the data comparison module; S2. According to the interaction scenario of the standardized test case, build a corresponding test scene in a real vehicle environment or a simulated real vehicle environment, and synchronously issue test instructions to the heavy truck human-computer interaction system to trigger human-computer interaction actions; S3. Start the data acquisition module, synchronously acquire image data, sound data, vibration data, and message data during the test process, and transmit them to the analysis processing module; the analysis processing module performs feature extraction, filtering processing, format analysis, and validity verification operations on the multi-source data, and outputs the processing results; S4. The data comparison module compares the processing results with the corresponding test cases in the standardized test case library one by one, determines whether they meet the preset standard, and generates comparison and determination results; S5. The data statistics module receives the comparison and determination results, outputs the test results including test pass or fail conclusions and optimization suggestions; after the single test process is completed, automatically generate statistical charts containing test pass rate and problem distribution; if there are items that do not pass, return to step S2 to rebuild the corresponding scene for retest, and if all test items pass, end the test.

[0011] Further, in step S1, the standardized test case includes four fields: scene number, test step, expected result, and determination threshold. The determination threshold includes the standardized format of instrument text, central control text, instrument icon, central control icon, sound, voice, and vibration parameters. Step S1 also includes adding, modifying, and deleting test cases according to the model upgrade and human-computer interaction function iteration requirements of the heavy truck, and updating the standardized test case library.

[0012] Further, in step S3, The analysis and processing module extracts the position of display elements, text information, and status indicators from image data using an image recognition algorithm. The image recognition algorithm combines the YOLO series algorithms with the SE attention mechanism to extract the position of display elements, text information, and status indicators from the image data. The analysis and processing module filters environmental noise from the sound and identifies the instruction content and prompt type through a sound recognition algorithm; the sound recognition algorithm removes environmental noise and extracts sound spectrum features through perceptual linear prediction (PLP) and identifies the speech instruction content and prompt type through a Conformer algorithm or a CNN algorithm. The analysis and processing module extracts intensity and frequency characteristics from vibration data and performs format parsing and validity verification on message data.

[0013] Furthermore, in step S5, if there are any failed items, the data statistics module marks the corresponding test case number, problem type, and data collection timestamp. During retesting, the original color technology and intermediate analysis results of the corresponding time period are retrieved first to help locate the cause of the problem.

[0014] Furthermore, the optimization suggestions in step S5 include the direction of test system adjustment and the direction of algorithm optimization.

[0015] As can be seen from the above technical solutions, the present invention has the following advantages: This invention discloses a testing system and method for human-machine interaction functions of a heavy-duty vehicle intelligent driving assistance system. The system simultaneously collects multi-source data, including images, sound, vibration, and messages, through a data acquisition module, improving anomaly detection rates and preventing observation omissions. The image recognition and sound recognition algorithms in the analysis and processing module shorten the testing cycle and meet the needs of batch iterative testing. A standardized test case library supports dynamic updates, adapting to heavy-duty vehicle model upgrades and function iterations without requiring system redevelopment, thus solving the problems of fixed and inflexible traditional test cases. The system supports both real-vehicle and simulated real-vehicle testing, covering complex scenarios such as extreme road conditions and cross-domain interactions, breaking through the limitations of existing testing scenarios. Simultaneously, data tracing and visualization of statistical charts enable rapid problem localization, and targeted optimization suggestions directly support the iteration of the human-machine interaction system, balancing testing accuracy, efficiency, and practicality, providing a reliable guarantee for the safe implementation of heavy-duty vehicle intelligent driving systems. Attached Figure Description

[0016] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the human-machine interaction function test system for the heavy-duty vehicle intelligent driving assistance system of this application.

[0018] Figure 2 This is a flowchart illustrating the human-machine interaction function test method for the heavy-duty vehicle intelligent driving assistance system of this application. Detailed Implementation

[0019] In the following detailed description of the human-machine interaction function testing system and method for intelligent driving assistance systems for heavy-duty vehicles, various embodiments of the present invention will be described more fully. The present invention may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present invention to the specific embodiments disclosed herein, but rather the present invention should be understood to cover all adjustments, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of the present invention.

[0020] It should be understood that, when used in this specification, the term "comprising" indicates the presence of the described feature, integral, step, operation, element, and / or component, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0021] The terms "one embodiment" or "some embodiments" used in this application mean that one or more embodiments of this application include the specific features, structures, or characteristics described in that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this application do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized.

[0022] To make the objectives, features, and advantages of this invention more apparent and understandable, specific embodiments and accompanying drawings will be used to clearly and completely describe the technical solutions protected by this invention. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0023] Please see Figure 1 The diagram shows a test system for the human-machine interaction function of a heavy-duty vehicle intelligent driving assistance system. The system includes: Data acquisition module, analysis and processing module, data comparison module, data statistics module; The data acquisition module is communicatively connected to the analysis and processing module. The data acquisition module is used to synchronously acquire multi-source data during the human-machine interaction process of heavy-duty vehicles. The multi-source data includes sensing data and message data. The sensing data includes image data, sound data and vibration data. The message data is the interactive command data transmitted by the vehicle system of the heavy-duty vehicle. The analysis and processing module is communicatively connected to the data comparison module. The analysis and processing module is used to receive multi-source data transmitted by the data acquisition module and perform processing operations on the multi-source data. The processing operations include feature extraction of image data, filtering and extraction of sound data, format parsing and validity verification of message data, and outputting the processing results. The data comparison module is communicatively connected to the data statistics module. The data comparison module has a pre-stored standardized test case library. The data comparison module is used to compare the processing results output by the analysis and processing module with the corresponding test cases in the standardized test case library one by one, determine whether the processing results meet the preset standards, and generate comparison and determination results. The data statistics module is used to receive the comparison and judgment results, output test results in real time including the conclusion of whether the test passed or failed and optimization suggestions, and automatically generate statistical charts including the test pass rate and the distribution of problems after a single test process is completed.

[0024] It should be noted that this embodiment achieves synchronous acquisition and automated analysis of multi-source interactive data (images, sounds, messages) through the interaction and collaboration of various modules, avoiding omissions caused by manual observation; it constructs a standardized test case library and automated comparison and statistics modules to shorten the testing cycle and improve batch testing efficiency; it supports dual testing scenarios of real vehicle and simulated real vehicle environments, adapts to different commercial heavy-duty vehicle models and interactive function iteration needs, and provides accurate suggestions for the optimization of intelligent driving assistance systems.

[0025] As a refinement and extension of the specific implementation methods described above, and to fully illustrate the specific implementation process in this embodiment, another human-machine interaction function testing system for a heavy-duty vehicle intelligent driving assistance system is provided. This system includes: Data acquisition module, analysis and processing module, data comparison module, data statistics module; The data acquisition module is communicatively connected to the analysis and processing module. The data acquisition module is used to synchronously acquire multi-source data during the human-machine interaction process of heavy-duty vehicles. The multi-source data includes sensing data and message data. The sensing data includes image data, sound data and vibration data. The message data is the interactive command data transmitted by the vehicle system of the heavy-duty vehicle. The data acquisition module includes an image acquisition unit, a sound acquisition unit, a message acquisition unit, and a vibration acquisition unit; The image acquisition unit is a high-definition industrial camera used to acquire image data, which is an image on the central control screen. The sound acquisition unit is a noise-canceling microphone array used to acquire sound data, which is system prompts or voice commands. The message acquisition unit includes a CAN bus interface and is used to acquire message data. The vibration acquisition unit is used to collect vibration data, which includes vibration information from the seat, seat belt, and steering wheel.

[0026] It should be noted that the high-definition industrial camera has a frame rate of no less than 30fps; the noise-reducing microphone array has a sampling rate of no less than 44.1kHz; and the message acquisition unit is connected to the heavy-duty vehicle CAN network through the CAN bus interface to acquire message data transmitted between the vehicle ECU and the vehicle human-machine interaction module.

[0027] The analysis and processing module is communicatively connected to the data comparison module. The analysis and processing module is used to receive multi-source data transmitted by the data acquisition module and perform processing operations on the multi-source data. The processing operations include feature extraction of image data, filtering and extraction of sound data, format parsing and validity verification of message data, and outputting the processing results. The analysis and processing module has built-in image recognition and sound recognition algorithms; The image recognition algorithm is used to extract the position of display elements, text information, and status indicators from image data; The sound recognition algorithm is used to filter out environmental noise in the sound and identify the content of the instruction and the type of prompt sound.

[0028] It should be noted that the image recognition algorithm is the YOLO series algorithm and the SE attention mechanism; the sound recognition algorithm is the Conformer algorithm or the CNN algorithm.

[0029] The data comparison module is communicatively connected to the data statistics module. The data comparison module has a pre-stored standardized test case library. The data comparison module is used to compare the processing results output by the analysis and processing module with the corresponding test cases in the standardized test case library one by one, determine whether the processing results meet the preset standards, and generate comparison and determination results. The standardized test case library in the data comparison module adds, modifies, and deletes test cases according to the needs of heavy-duty vehicle model upgrades and human-computer interaction function iterations. Each test case includes four fields: scenario number, test steps, expected result, and judgment threshold. The determination threshold is in a standardized format, including instrument text, central control text, instrument icon, central control icon, sound, voice, and vibration parameters.

[0030] In some embodiments, examples of the standardized format of test cases in the standardized test case library are shown in Table 1: Table 1. Examples of standardized test case formats

[0031] The data statistics module is used to receive the comparison and judgment results, output test results in real time including the conclusion of whether the test passed or failed and optimization suggestions, and automatically generate statistical charts including the test pass rate and the distribution of problems after a single test process is completed.

[0032] The statistical charts output by the data statistics module include bar charts, line charts, and pie charts; The bar chart is used to show the number of tests that passed and failed in different interaction scenarios; The line graph is used to show the trend of test results for the same scenario in multiple retests; The pie chart is used to show the percentage distribution of various test question types.

[0033] Please see Figure 2 The diagram shows a flowchart of a testing method for the human-machine interaction function of an intelligent driving assistance system for heavy-duty vehicles. The method includes the following steps: S1. Establish a real-time data acquisition link adapted to the heavy-duty vehicle's onboard system, enabling the data acquisition module to stably access the heavy-duty vehicle's data transmission channel; based on the testing requirements for the heavy-duty vehicle's human-machine interaction function, write standardized test cases covering interactive scenarios such as button operation response, voice command recognition, display interface feedback, and vibration feedback, and store the standardized test cases in the standardized test case library of the data comparison module; S2. Based on the interaction scenarios of the standardized test cases, construct corresponding test scenarios in a real vehicle environment or a simulated real vehicle environment, and synchronously issue test instructions to the heavy-duty vehicle human-machine interaction system to trigger human-machine interaction actions; S3. Start the data acquisition module to synchronously acquire image data, sound data, vibration data, and message data during the test process, and transmit them to the analysis and processing module; the analysis and processing module performs feature extraction, filtering, format parsing, and validity verification on the multi-source data, and outputs the processing results. S4. The data comparison module compares the processing results with the corresponding test cases in the standardized test case library one by one to determine whether they meet the preset standards and generates the comparison judgment results. S5. The data statistics module receives the comparison and judgment results and outputs test results including the conclusion of whether the test passed or failed and optimization suggestions. After a single test process is completed, it automatically generates statistical charts containing the test pass rate and the distribution of problems. If there are any failed items, it returns to step S2 to reconstruct the corresponding scenario for retesting. The test ends when all test items pass.

[0034] It should be noted that the real-time data acquisition link established in step S1 adopts 5G or Ethernet transmission mode, and the data transmission rate is not less than 100Mbps; the image acquisition unit frame rate of the data acquisition module in step S3 is not less than 30fps, the sound acquisition unit sampling rate is not less than 44.1kHz, and the timestamp deviation of multi-source data acquisition does not exceed 50ms to ensure data synchronization and correlation.

[0035] As a refinement and extension of the specific implementation of the above embodiments, in order to fully illustrate the specific implementation process of this embodiment, another method for testing the human-machine interaction function of a heavy-duty vehicle intelligent driving assistance system is provided. This method includes the following steps: S1. Establish a real-time data acquisition link adapted to the heavy-duty vehicle's onboard system, enabling the data acquisition module to stably access the heavy-duty vehicle's data transmission channel; based on the testing requirements for the heavy-duty vehicle's human-machine interaction function, write standardized test cases covering interactive scenarios such as button operation response, voice command recognition, display interface feedback, and vibration feedback, and store the standardized test cases in the standardized test case library of the data comparison module; In step S1, the standardized test case includes four fields: scenario number, test steps, expected results, and judgment threshold. The determination thresholds include standardized formats for instrument text, central control text, instrument icons, central control icons, sound, voice, and vibration parameters. Step S1 also includes adding, modifying, and deleting test cases and updating the standardized test case library based on the needs of heavy-duty vehicle model upgrades and human-machine interaction function iterations.

[0036] S2. Based on the interaction scenarios of the standardized test cases, construct corresponding test scenarios in a real vehicle environment or a simulated real vehicle environment, and synchronously issue test instructions to the heavy-duty vehicle human-machine interaction system to trigger human-machine interaction actions; S3. Start the data acquisition module to synchronously acquire image data, sound data, vibration data, and message data during the test process, and transmit them to the analysis and processing module; the analysis and processing module performs feature extraction, filtering, format parsing, and validity verification on the multi-source data, and outputs the processing results; in step S3: The analysis and processing module extracts the position of display elements, text information, and status indicators from image data using an image recognition algorithm. The image recognition algorithm combines the YOLO series algorithms with the SE attention mechanism to extract the position of display elements, text information, and status indicators from the image data. The analysis and processing module filters environmental noise from the sound and identifies the instruction content and prompt type through a sound recognition algorithm; the sound recognition algorithm removes environmental noise and extracts sound spectrum features through perceptual linear prediction (PLP) and identifies the speech instruction content and prompt type through a Conformer algorithm or a CNN algorithm. The analysis and processing module extracts intensity and frequency characteristics from vibration data and performs format parsing and validity verification on message data.

[0037] In some embodiments, taking the scenario of a driver triggering cruise control to activate human-computer interaction while a heavy vehicle is traveling at high speed as an example: After the data acquisition module is started synchronously, the high-definition industrial camera (30fps) acquires dynamic image data of the central control screen display interface and the instrument panel, the noise reduction microphone array (44.1kHz sampling rate) acquires the driver's "activate cruise control" voice command and system feedback prompts, the vibration sensing unit acquires steering wheel vibration feedback data, and the message acquisition unit obtains the transmission messages between the vehicle ECU and the human-machine interaction module through the CAN bus interface. The four types of data are synchronously transmitted to the analysis and processing module via the 5G link, with the timestamp deviation controlled within 50ms.

[0038] First, the YOLO algorithm is used to construct a multi-scale feature pyramid for the image, quickly detecting target objects such as text areas on the central control screen and icons on the instrument panel, and outputting the coordinates of the target bounding boxes. Then, the SE attention mechanism module is connected to adaptively adjust the channel weights of the feature maps extracted by YOLO, enhance the response intensity of key features such as text pixels and icon colors, suppress background interference, and finally accurately extract the central control text "Cruise is on", the blue activation icon on the instrument panel, and the coordinates of the icon display position corresponding to cruise control activation, forming structured image feature data.

[0039] The collected sound signal is preprocessed using the Perceptual Linear Prediction (PLP) algorithm. The spectrum is nonlinearly compressed according to the characteristics of human hearing to remove environmental interference such as high-speed wind noise and engine noise, and to extract clean sound spectrum features. In the second step, the spectrum features are input into the Conformer algorithm and the CNN algorithm respectively. The Conformer algorithm captures the temporal dependency of the voice command through a multi-head attention mechanism and identifies the voice content for activating cruise control. The CNN algorithm extracts the spectrum texture features of the prompt tone through convolutional layers and determines that the system feedback tone is the type of cruise control activation success prompt tone. The results of the two algorithms are cross-validated and the sound recognition result is output.

[0040] Vibration data processing is performed using a feature extraction algorithm. Time-domain analysis is conducted on the collected steering wheel vibration signals to calculate the peak vibration intensity (unit: m / s²). 2 The vibration characteristic parameter set is formed by combining the frequency (unit: Hz). The message data processing parses the data frame according to the CAN bus protocol format, extracts key fields such as the interaction command ID and execution status code, and completes the message validity verification by checking the CRC value of the verification field and the data length, and eliminates abnormal transmission data.

[0041] The analysis and processing module integrates image structured features, sound recognition results, vibration feature parameters, and valid message data into a unified format processing result, which is then synchronously transmitted to the data comparison module.

[0042] S4. The data comparison module compares the processing results with the corresponding test cases in the standardized test case library one by one to determine whether they meet the preset standards and generates the comparison judgment results. S5. The data statistics module receives the comparison and judgment results and outputs test results including the conclusion of whether the test passed or failed and optimization suggestions. After a single test process is completed, it automatically generates statistical charts containing the test pass rate and the distribution of problems. If there are any failed items, it returns to step S2 to reconstruct the corresponding scenario for retesting. The test ends when all test items pass.

[0043] In step S5, if there are any failed items, the data statistics module marks the corresponding test case number, problem type, and data collection timestamp. During retesting, the original color technology and intermediate analysis results of the corresponding time period are retrieved first to help locate the cause of the problem.

[0044] The optimization suggestions mentioned in step S5 include the direction of test system adjustment and the direction of algorithm optimization.

[0045] It should be noted that the adjustment directions of the aforementioned testing system include: Hardware adaptation and adjustment of the data acquisition module: adjust the installation angle, focal length or frame rate of the image acquisition unit, optimize the pickup angle, installation position or sampling rate of the sound acquisition unit, adjust the installation position or number of vibration acquisition units, and adapt to the CAN bus interface communication protocol of the message acquisition unit or increase interface redundancy. Data transmission links are optimized and adjusted by upgrading transmission bandwidth, optimizing network signal reception, adjusting transmission protocols, or adding local caching modules to ensure multi-source data synchronization and transmission stability. Modular assembly and adjustment allow for the individual or combined use of each acquisition unit according to specific testing needs, adapting to the message acquisition requirements of multi-domain ECUs or adjusting the module communication interface to adapt to the protocol. The standardized test case library has been iteratively adjusted, with the addition of test cases for extreme scenarios or new features, and the judgment thresholds of existing test cases have been optimized. The optimization directions of the algorithm include: Image recognition algorithm optimization: Adjust the anchor box parameters of YOLO series algorithms and the channel weights of SE attention mechanism; add image preprocessing algorithms or expand training samples for multiple vehicle models. Optimize the voice recognition algorithm by improving the parameters of the PLP filtering algorithm, optimizing the attention mechanism window of the Conformer algorithm, the convolution kernel size of the CNN algorithm, or supplementing training samples with specific accents; Data verification and fusion algorithm optimization, adding multi-field cross-verification logic for messages, adding multi-source data fusion algorithm or optimizing vibration data feature extraction algorithm; The dynamic calibration algorithm for the judgment threshold is optimized by introducing a machine learning algorithm to automatically adjust the judgment threshold and establishing a threshold optimization model based on historical test data.

[0046] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0047] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A heavy truck intelligent driving assistance system human-computer interaction function test system, characterized in that, The system comprises: a data acquisition module, an analysis processing module, a data comparison module, and a data statistics module; the data acquisition module is in communication connection with the analysis processing module, the data acquisition module is used for synchronously acquiring multi-source data in the human-machine interaction process of a heavy truck, the multi-source data comprises perception data and message data, the perception data comprises image data, sound data, and vibration data, and the message data is interactive instruction data transmitted by a vehicle-mounted system of the heavy truck; the analysis processing module is in communication connection with the data comparison module, the analysis processing module is used for receiving the multi-source data transmitted by the data acquisition module and performing processing operations on the multi-source data, the processing operations comprise feature extraction on the image data, filtering processing and extraction on the sound data, and format analysis and validity checking on the message data, and the processing results are outputted; the data comparison module is in communication connection with the data statistics module, the data comparison module pre-stores a standardized test case library, the data comparison module is used for comparing the processing results outputted by the analysis processing module with corresponding test cases in the standardized test case library one by one, determining whether the processing results meet preset standards, and generating comparison and determination results; the data statistics module is used for receiving the comparison and determination results, outputting test results in real time, and outputting test results including test passing or failing and optimization suggestions, and automatically generating statistical charts and tables including test passing rates and problem distribution situations after a single test process is completed.

2. The heavy-duty vehicle intelligent driving assistance system human-machine interaction function test system according to claim 1, characterized in that, The data acquisition module comprises an image acquisition unit, a sound acquisition unit, a message acquisition unit, and a vibration acquisition unit; the image acquisition unit is a high-definition industrial camera, used for acquiring image data, and the image data is a central control screen image; the sound acquisition unit is a noise reduction microphone array, used for acquiring sound data, and the sound data is a system prompt tone or a voice instruction; the message acquisition unit comprises a CAN bus interface, and the message acquisition unit is used for acquiring message data; the vibration acquisition unit is used for acquiring vibration data, and the vibration data comprises vibration information of a seat, a safety belt, and a steering wheel.

3. The heavy-duty vehicle intelligent driver assistance system human-machine interaction function test system according to claim 1, characterized in that, The analysis processing module is built-in with an image recognition algorithm and a sound recognition algorithm; the image recognition algorithm is used for extracting display element positions, text information, and state identifiers in the image data; the sound recognition algorithm is used for filtering environmental noise in the sound and recognizing instruction contents and prompt tone types.

4. The heavy-duty vehicle intelligent driver assistance system human-machine interaction function test system according to claim 1, characterized in that, The standardized test case library in the data comparison module is added, modified, and deleted according to vehicle model upgrades and human-machine interaction function iteration requirements of the heavy truck; each test case comprises four fields of scene number, test steps, expected results, and determination threshold; the determination threshold is in a standardized format, comprising instrument text, central control text, instrument icons, central control icons, sound, voice, and vibration parameters.

5. The heavy-duty vehicle intelligent driver assistance system human-machine interaction function test system according to claim 1, characterized in that, The statistical charts and tables outputted by the data statistics module comprise column charts, line charts, and pie charts; the column chart is used for displaying the number of test passing and the number of test failing in different interaction scenes; the line chart is used for displaying the test result change trend of the same scene in multiple retests; the pie chart is used for displaying the proportion of test passing and test failing in different interaction scenes. The pie chart is used to show the proportion distribution of various types of test problems.

6. A method for testing human-machine interaction function of intelligent driving assistance system for heavy-duty vehicles, the method is based on the human-machine interaction function test system according to any one of claims 1-5, characterized in that, The method comprises the following steps: S1. Build a real-time data acquisition link compatible with the on-board system of a heavy-duty vehicle, so that the data acquisition module can be stably connected to the data transmission channel of the heavy-duty vehicle; according to the test requirements of the human-machine interaction function of the heavy-duty vehicle, write standardized test cases covering key operation response, voice command recognition, display interface feedback, and vibration feedback of the interactive scene, and store the standardized test cases in the standardized test case library of the data comparison module; S2. According to the interactive scene of the standardized test case, build a corresponding test scene in a real vehicle environment or a simulated real vehicle environment, and synchronously issue test instructions to the heavy-duty vehicle human-machine interaction system to trigger human-machine interaction actions; S3. Start the data acquisition module to synchronously collect image data, sound data, vibration data, and message data during the test process, and transmit them to the analysis processing module; the analysis processing module performs feature extraction, filtering processing, format analysis, and validity verification operations on the multi-source data, and outputs the processing results; S4. The data comparison module compares the processing results with the corresponding test cases in the standardized test case library one by one to determine whether they meet the preset standards and generates comparison and determination results; S5. The data statistics module receives the comparison and determination results, outputs test results including test pass or fail conclusions and optimization suggestions, and automatically generates statistical charts and tables containing test pass rates and problem distribution situations after a single test process is completed; if there are items that do not pass, return to step S2 to rebuild the corresponding scene for retesting, and if all test items pass, end the test.

7. The heavy-duty vehicle intelligent driver assistance system human-machine interaction function test method according to claim 6, characterized in that, In step S1, the standardized test case includes four fields: scene number, test steps, expected results, and determination threshold; The determination threshold includes the standardized format of instrument text, central control text, instrument icon, central control icon, sound, voice, and vibration parameters; Step S1 further includes adding, modifying, and deleting test cases according to the vehicle model upgrade and human-machine interaction function iteration requirements of the heavy-duty vehicle, and updating the standardized test case library.

8. The heavy-duty vehicle intelligent driver assistance system human-machine interaction function test method according to claim 6, characterized in that, In step S3: The analysis processing module extracts the display element position, text information, and state identifier in the image data through an image recognition algorithm; the image recognition algorithm uses a combination of YOLO series algorithm and SE attention mechanism to extract the display element position, text information, and state identifier in the image data; The analysis processing module filters environmental noise in the sound and recognizes the instruction content and prompt tone type through a sound recognition algorithm; the sound recognition algorithm removes environmental noise and extracts sound spectrum features through a perceptual linear prediction (PLP) method, and recognizes voice instruction content and prompt tone type through a conformer algorithm or CNN algorithm; The analysis processing module extracts intensity and frequency characteristics from the vibration data, and performs format analysis and validity verification on the message data.

9. The heavy-duty vehicle intelligent driver assistance system human-machine interaction function test method according to claim 6, characterized in that, In step S5, if there are items that do not pass, the data statistics module marks the corresponding test case number, problem type, and data acquisition timestamp, and when retesting, preferentially retrieves the original color technology and analysis processing intermediate results of the corresponding time period to assist in locating the problem cause.

10. The heavy-duty vehicle intelligent driver assistance system human-machine interaction function test method according to claim 6, characterized in that, The optimization suggestion in step S5 includes a test system adjustment direction and an algorithm optimization direction.

Images