A simulation test method and device for intelligent driving cockpit

Abstract

The present application relates to the field of vehicle testing technology, and in particular to a simulation test method and device for an intelligent driving cockpit, wherein a simulation environment is configured for the driving cockpit of a vehicle to be tested, and a training set that conforms to the driving cockpit is obtained; a camera is used to collect data on the internal scene of the driving cockpit to obtain an internal scene image of the driving cockpit; the training set is added to the internal scene image and then input into a test model, and the training set is identified to obtain a recognition result; based on the recognition result, a response is made, the internal scene of the vehicle to be tested is controlled, and the vehicle to be tested is verified and optimized. The simulation platform of the present invention can simulate various complex and extreme driving scenarios, as well as lighting environments, provide comprehensive test coverage, generate training sets in a virtual environment, test the performance of the internal system of the intelligent driving cockpit under these intrusions, and continuously optimize and improve the system to improve its resistance to intrusions and overall safety.

Description

Technical Field

[0001] The present application relates to the field of vehicle testing technology, and in particular to a simulation testing method and device for an intelligent driving cockpit. Background Art

[0002] Testing in a physical environment is an ideal method for evaluating the performance of intelligent driving systems, but this method has the following limitations: High cost and time-consuming: Large-scale testing in a real test environment requires a lot of time and money. Each test involves huge resource consumption in terms of venues, vehicles, manpower, etc.; Safety risks: Testing the intrusion of training sets in a real environment may cause safety accidents and threaten the safety of testers and the public; Limited scenario coverage: Even the most comprehensive physical tests cannot cover all possible driving scenarios and conditions, especially some extreme cases.

[0003] Using datasets for research is another common approach, but it also has significant limitations:

[0004] Theorization and abstraction: Data sets usually cannot fully reflect the complexity of the real world. The uncertainty and variability in the real environment are difficult to capture in static data sets; Lack of dynamics: Most data sets are static and cannot simulate dynamic interactions and real-time changing environments. This makes the research results invalid in practical applications; Generation and testing of training sets: There are large theoretical assumptions in generating and testing training sets in data sets, and the actual results may be far from the performance in the real world. Summary of the invention

[0005] To achieve the above objectives, this application provides the following technical solutions:

[0006] According to a first aspect of the present invention, the present invention claims protection for a simulation test method for an intelligent driving cockpit, comprising:

[0007] Performing simulation environment configuration on the cockpit of the vehicle to be tested to obtain a training set that conforms to the cockpit;

[0008] Using a camera to collect data about the interior scene of the cockpit to obtain an image of the interior scene of the cockpit;

[0009] Adding the training set to the internal scene image and then inputting it into a test model, recognizing the training set to obtain a recognition result;

[0010] In response, the interior scene of the vehicle to be tested is controlled and verification and optimization of the vehicle to be tested is performed based on the recognition result.

[0011] Furthermore, the configuration of the simulation environment for the cockpit of the vehicle to be tested also includes:

[0012] Configure the simulation environment for the cockpit of the test vehicle and use the simulation platform to create a virtual cockpit interior scene;

[0013] The cockpit interior scene at least includes:

[0014] Instrument panel, infotainment system, seat, door sensor, temperature sensor, live body legacy scene;

[0015] Setting environmental parameters of the interior scene of the cockpit to simulate real driving conditions;

[0016] The environmental parameters include at least:

[0017] Light and sound.

[0018] Furthermore, the obtaining of a training set that conforms to the cockpit further includes:

[0019] Generate a training set for the cockpit internal system using a training set generation algorithm;

[0020] The classification of the training set at least includes visual interference class, sound interference class, and touch screen misoperation class;

[0021] The training set is dynamically generated and applied so that the test meets actual driving conditions.

[0022] Furthermore, the method of collecting data of the interior scene of the cockpit using a camera to obtain an image of the interior scene of the cockpit also includes:

[0023] When collecting data on the interior scene of the cockpit, a target object of the interior scene and a valid position of the target object are marked.

[0024] Furthermore, the adding of the training set to the internal scene image and then inputting the test model to recognize the training set to obtain the recognition result also includes:

[0025] Adding the training set to the target object of the internal scene and the valid position of the target object to form an internal scene training image;

[0026] The internal scene training image is input into the test model through a program interface, and the training set is loaded into the effective position of the target object in the preset position to realize a dynamic intrusion.

[0027] Furthermore, the responding based on the recognition result, controlling the internal scene of the vehicle to be tested and verifying and optimizing the vehicle to be tested also includes:

[0028] Apply the generated recognition results in a simulation environment to test the reaction and decision of the vehicle to be tested under the intrusion environment, and record the reaction performance of the vehicle to be tested under different intrusions;

[0029] The reaction performance includes at least:

[0030] Error rate, response time, system stability.

[0031] Further, the method also includes:

[0032] The target object library, training set database and driving scenario database are preset;

[0033] The preset target object in the target object library is the object to be invaded;

[0034] The target object library at least includes:

[0035] Driver status, environment status, Face ID.

[0036] Furthermore, when the cockpit interior scene is a dashboard, it also includes:

[0037] After the vehicle to be tested enters the tunnel, the ambient light changes in the interior scene of the cockpit are simulated, the brightness of the dashboard of the driving cockpit is automatically adjusted, the daytime running lights are automatically turned on, and the changes in the vehicle are observed after being added to the training set.

[0038] Furthermore, when the cabin interior scene is a scene left behind by a living body, it also includes:

[0039] The living subject is a child or a pet;

[0040] Simulating the activities of children or pets in the vehicle to be tested in different positions and states, including at least being still, moving, and making sounds;

[0041] When a child is left behind, simulate the scenario of the child being left in the car, including at least moving, crying, and playing in the seat;

[0042] When left behind for a pet, simulate the pet being left in the car, including at least moving, barking, and attempting to escape;

[0043] Generate a training set for interfering with a camera inside a cockpit, wherein the training set is used to disrupt the camera's recognition of children or pets;

[0044] Using a sound training set generation algorithm, generating a sound training set for interfering with a microphone inside the cabin, wherein the sound training set is used to interfere with the microphone's detection of a child's crying or a pet's barking;

[0045] The sound training set includes noises of different frequencies and volumes to simulate interference factors in a real environment;

[0046] The temperature training set generation technology is used to simulate the interference of the temperature sensor in the car, and a training set is generated to make the temperature sensor unable to accurately sense the temperature changes inside the cabin, resulting in the failure to detect the high temperature environment in time. The temperature training set is used to simulate the failure of the temperature sensor or human intervention, affecting the system's response to high temperature.

[0047] According to a second aspect of the present invention, the present invention claims protection for a simulation test device for an intelligent driving cockpit, comprising:

[0048] one or more processors;

[0049] A memory having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the simulation test method for the intelligent driving cockpit.

[0050] The present application relates to the field of vehicle testing technology, and in particular to a simulation test method and device for an intelligent driving cockpit, wherein a simulation environment is configured for the driving cockpit of a vehicle to be tested, and a training set that conforms to the driving cockpit is obtained; a camera is used to collect data on the internal scene of the driving cockpit to obtain an internal scene image of the driving cockpit; the training set is added to the internal scene image and then input into a test model, and the training set is identified to obtain a recognition result; based on the recognition result, a response is made, the internal scene of the vehicle to be tested is controlled, and the vehicle to be tested is verified and optimized. The simulation platform of the present invention can simulate various complex and extreme driving scenarios, as well as lighting environments, provide comprehensive test coverage, generate training sets in a virtual environment, test the performance of the internal system of the intelligent driving cockpit under these intrusions, and continuously optimize and improve the system to improve its resistance to intrusions and overall safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A workflow diagram of a simulation test method for an intelligent driving cockpit as claimed in an embodiment of the present application;

[0052] Figure 2 This is a simulation example diagram of a simulation test method for an intelligent driving cockpit as claimed in an embodiment of the present application. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] The terms "first", "second" and "third" in this application are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined as "first", "second" and "third" can explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. All directional indications (such as up, down, left, right, front, back...) in the embodiments of this application are only used to explain the relative position relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication also changes accordingly. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices.

[0055] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0056] According to the first embodiment of the present invention, the present invention claims a simulation test method for an intelligent driving cockpit, referring to Figure 1 ,include:

[0057] Performing simulation environment configuration on the cockpit of the vehicle to be tested to obtain a training set that conforms to the cockpit;

[0058] Using a camera to collect data about the interior scene of the cockpit to obtain an image of the interior scene of the cockpit;

[0059] Adding the training set to the internal scene image and then inputting it into a test model, recognizing the training set to obtain a recognition result;

[0060] In response, the interior scene of the vehicle to be tested is controlled and verification and optimization of the vehicle to be tested is performed based on the recognition result.

[0061] Furthermore, the configuration of the simulation environment for the cockpit of the vehicle to be tested also includes:

[0062] Configure the simulation environment for the cockpit of the test vehicle and use the simulation platform to create a virtual cockpit interior scene;

[0063] The cockpit interior scene at least includes:

[0064] Instrument panel, infotainment system, seat, door sensor, temperature sensor, live body legacy scene;

[0065] Setting environmental parameters of the interior scene of the cockpit to simulate real driving conditions;

[0066] The environmental parameters include at least:

[0067] Light and sound.

[0068] In this embodiment, it also includes:

[0069] The instrument panel is used to display important information such as the vehicle's speed, rotation speed, and fuel level. In the simulation test, the specific application target of the instrument panel is to accurately present the vehicle's operating status information so that the driver can understand the vehicle's status in a timely manner. For example, during the simulated vehicle driving process, the instrument panel should be able to accurately display the speed change within the range of 0-120 km / h, and the process of the fuel level from full to gradually decreasing.

[0070] The infotainment system includes functions such as navigation, audio, and Bluetooth connection. Its application target is to provide convenient entertainment and information services. For example, navigation should be able to accurately plan routes and update road conditions in real time, and audio should be able to clearly play various audio files. In terms of parameter indicators, for example, the positioning accuracy of navigation should be within a few meters, and the volume of the audio can be adjusted from 0 to 100%.

[0071] The seat is concerned with comfort and safety. The specific application target is to provide good support and protection for the driver and passengers. In the collision simulation, the seat should be able to effectively reduce the injury of the occupants. The seat parameter indicators include the adjustable range of the seat tilt angle from 0 to 30 degrees, and the seat material should have a certain degree of softness and breathability.

[0072] The door sensor is used to detect the open and closed state of the door, ensuring that the door is firmly closed and ensuring driving safety. Its application target is to promptly feedback the door status information to the driver, such as issuing an alarm when the door is not closed properly. The sensitivity of the door sensor should be able to accurately detect the slight opening state of the door.

[0073] The temperature sensor is used to monitor the temperature inside the cabin, provide a control basis for the air conditioning system, and ensure a comfortable temperature inside the car. The specific application target is to accurately sense the temperature changes inside the car so that the air conditioning system can automatically adjust the temperature. The measurement accuracy of the temperature sensor should be within the range of ±1 degree Celsius.

[0074] The live body left behind scenario mainly simulates the situation where a child or pet is left behind in the car. The application target object is to detect whether there is a live body left in the car and issue an alarm in time. For example, when a child or pet is detected to be stationary in the car for more than a certain period of time (such as 10 minutes) or makes an abnormal sound (such as crying, barking), the system should be able to issue an alarm to remind the driver.

[0075] Set the environmental parameters of the interior scene of the cockpit to simulate real driving conditions. The environmental parameters include at least lighting and sound. For example, the lighting parameters can simulate the light intensity and color temperature in different scenes such as daytime, nighttime, and tunnels. The light intensity during the day can be 1000-5000 lux, 10-50 lux at night, and 100-500 lux in tunnels. The sound parameters can simulate the engine noise, wind noise, tire noise, etc. during the vehicle's driving process, as well as the sounds of the external environment, such as rain, horn, etc. The volume of the sound can be adjusted between 50-80 decibels.

[0076] Reference Figure 2 ,On the simulation platform, the internal scene image is input into the internal test model of the cockpit, the internal environment of the cockpit is controlled by the virtual controller, and the lighting angle is adjusted to set the simulation environment.

[0077] Furthermore, the obtaining of a training set that conforms to the cockpit further includes:

[0078] Generate a training set for the cockpit internal system using a training set generation algorithm;

[0079] The classification of the training set at least includes visual interference class, sound interference class, and touch screen misoperation class;

[0080] The training set is dynamically generated and applied so that the test meets actual driving conditions.

[0081] In this embodiment, it also includes:

[0082] Visual distractions include:

[0083] Bright Light Flash: Simulates sudden bright light, such as the rapid flickering of a flashlight.

[0084] Fog Occlusion: Generates a fog-like image that blocks part or all of the view.

[0085] Color interference: Presenting strongly contrasting or confusing colors that interfere with the recognition of important information.

[0086] False signs: False warning signs or messages are displayed on the instrument panel or display.

[0087] Sound disturbances include:

[0088] Sharp noises: such as high-pitched screams or piercing sirens.

[0089] Noisy background sound: simulates the noisy sounds on the street, the roar of machines, etc.

[0090] Voice Interference: Plays voice messages that are similar to or confusing with driving instructions.

[0091] Sudden loud noises: such as explosions or crashes, which distract the driver from paying attention.

[0092] Touch screen misoperation categories include:

[0093] Random touch: randomly generate touch points on the touch screen to simulate accidental touch operations.

[0094] Multi-point simultaneous touch: Multiple touch points are triggered at the same time, causing the system to misunderstand the instructions.

[0095] Sliding interference: fast or irregular sliding operations affect normal interface switching.

[0096] Repeated clicks: Clicking the same button quickly and continuously may cause the system to respond incorrectly.

[0097] Furthermore, the method of collecting data of the interior scene of the cockpit using a camera to obtain an image of the interior scene of the cockpit also includes:

[0098] When collecting data on the interior scene of the cockpit, a target object of the interior scene and a valid position of the target object are marked.

[0099] Furthermore, the adding of the training set to the internal scene image and then inputting the test model to recognize the training set to obtain the recognition result also includes:

[0100] Adding the training set to the target object of the internal scene and the valid position of the target object to form an internal scene training image;

[0101] The internal scene training image is input into the test model through a program interface, and the training set is loaded into the effective position of the target object in the preset position to realize a dynamic intrusion.

[0102] In this embodiment, it also includes:

[0103] In sensor interference intrusion scenarios, for example, emitting strong light to the camera will cause the camera to be overexposed, making the captured image seriously distorted and blurred, and unable to provide valuable materials for subsequent analysis and processing. Using lasers or high temperatures to damage the camera is a more direct and destructive intrusion method, which may permanently damage the key components of the camera and completely make it unable to work normally.

[0104] Strong light interference can also have a serious impact on lidar or ultrasonic sensors, causing huge errors in measuring distance, detecting objects, etc., affecting the normal operation of systems that rely on these sensors.

[0105] In the case of a lost child or pet:

[0106] For example, in a child or pet lost scenario, the target object is a child or a pet. When simulating this scenario, first determine the effective position of the child or pet in the car, for example, the child may be active on the seat, and the pet may move on the floor in the car. For the addition of visual interference training sets, some interference images can be generated, such as simulating shadows or reflections caused by light changes, so that they cover the effective position of the child or pet to disrupt the camera's recognition of the child or pet. For example, when a child sits on a seat, some blurred light and shadow or color spots can be added to the area, making it difficult for the camera to clearly capture the child's figure. For the addition of sound interference training sets, noise of a specific frequency and volume can be played, such as playing noises similar to those emitted by other devices in the car or noisy sounds in the external environment near the effective position of the child or pet, to interfere with the microphone's detection of the child's crying or the pet's calling. For example, when a pet is on the floor in the car, sounds similar to wind, rain or other vehicles driving can be played near the area to cover up the possible calls of the pet.

[0107] Furthermore, the responding based on the recognition result, controlling the internal scene of the vehicle to be tested and verifying and optimizing the vehicle to be tested also includes:

[0108] Apply the generated recognition results in a simulation environment to test the reaction and decision of the vehicle to be tested under the intrusion environment, and record the reaction performance of the vehicle to be tested under different intrusions;

[0109] The reaction performance includes at least:

[0110] Error rate, response time, system stability.

[0111] In this embodiment, it also includes:

[0112] Error rate scenario:

[0113] In the live body left behind scenario, if the system fails to detect the child or pet left behind in time, or issues an alarm incorrectly, the ratio of such occurrences to the total number of tests is counted. For example, in 50 live body left behind scenario tests, if the system fails to detect the child or pet left behind 5 times, the error rate is 10%.

[0114] Reaction time Reaction performance:

[0115] In the driver detection scenario, the time it takes for the driver detection to identify driver abnormalities, such as 0.3 seconds, is recorded when the driver loses the ability to control the vehicle due to fatigue, illness or other unexpected reasons.

[0116] System stability response performance:

[0117] For the infotainment system, when it is disturbed by the sound interference training set, observe whether the system can continue to operate normally, such as whether the music playback is interrupted, whether the navigation can navigate accurately, etc. If the number of abnormalities within a certain period of time (such as 1 hour) is small, it means that the system stability is good.

[0118] Further, the method also includes:

[0119] The target object library, training set database and driving scenario database are preset;

[0120] The preset target object in the target object library is the object to be invaded;

[0121] The target object library at least includes:

[0122] Driver status, environment status, Face ID.

[0123] In this embodiment, it also includes:

[0124] Driver status:

[0125] Scenario: Fatigue driving monitoring.

[0126] Target objects: driver’s facial expression, eye closure status, head posture, etc.

[0127] Specific effective position (effective area): The camera focuses on the driver's face, especially the eyes and head area, to detect the degree of eye closure, blinking frequency, and head tilt angle, etc.

[0128] Environmental status:

[0129] Scenario: Temperature adjustment in the car.

[0130] Target objects: Temperature sensors, window positions, air conditioning systems, etc.

[0131] Specific effective position (effective area): Temperature sensors are distributed in various locations in the car, such as near the seats, above the center console, etc., to accurately sense the temperature of different areas in the car; the effective position of the window lies in its switch status to determine whether ventilation is needed; the air outlet and control panel of the air conditioning system are also important effective areas.

[0132] Face ID:

[0133] Scenario: Identity recognition and personalization.

[0134] Target object: Driver’s facial features.

[0135] Specific effective position (effective area): The camera is aimed at the driver's face to obtain facial feature information for identification of the driver, thereby adjusting personalized settings such as seat position, rearview mirror angle, music preference, etc. For example, when the driver gets in the car, the camera is located above the steering wheel or near the rearview mirror inside the car, and can clearly capture the driver's face.

[0136] Furthermore, when the cockpit interior scene is a dashboard, it also includes:

[0137] After the vehicle to be tested enters the tunnel, the ambient light changes in the interior scene of the cockpit are simulated, the brightness of the dashboard of the driving cockpit is automatically adjusted, the daytime running lights are automatically turned on, and the changes in the vehicle are observed after being added to the training set.

[0138] Furthermore, when the cabin interior scene is a scene left behind by a living body, it also includes:

[0139] The living subject is a child or a pet;

[0140] Simulating the activities of children or pets in the vehicle to be tested in different positions and states, including at least being still, moving, and making sounds;

[0141] When a child is left behind, simulate the scenario of the child being left in the car, including at least moving, crying, and playing in the seat;

[0142] When left behind for a pet, simulate the pet being left in the car, including at least moving, barking, and attempting to escape;

[0143] Generate a training set for interfering with a camera inside a cockpit, wherein the training set is used to disrupt the camera's recognition of children or pets;

[0144] Using a sound training set generation algorithm, generating a sound training set for interfering with a microphone inside the cabin, wherein the sound training set is used to interfere with the microphone's detection of a child's crying or a pet's barking;

[0145] The sound training set includes noises of different frequencies and volumes to simulate interference factors in a real environment;

[0146] The temperature training set generation technology is used to simulate the interference of the temperature sensor in the car, and a training set is generated to make the temperature sensor unable to accurately sense the temperature changes inside the cabin, resulting in the failure to detect the high temperature environment in time. The temperature training set is used to simulate the failure of the temperature sensor or human intervention, affecting the system's response to high temperature.

[0147] Among them, in this embodiment, the test results are analyzed, the weaknesses and vulnerabilities of the system are identified, and based on the analysis results, the AI ​​model is adjusted and optimized, and an effective defense strategy is designed to improve the robustness of the system.

[0148] System Verification:

[0149] Test again in a simulation environment to verify the performance of the optimized system under the intrusion of the training set, ensuring that the system can operate stably under a variety of complex and extreme conditions.

[0150] Training set invasion:

[0151] The training set is dynamically generated, and the embodiment of the present application does not limit the number of training sets generated in the same period or different periods of the same road section, the type of intrusion, the intrusion object, etc. In some implementations, the embodiment of the present application can add target objects from the preset target object library to the simulation test scene according to the current test environment, and perform disturbance processing on the added target objects to obtain the training set.

[0152] Analyze the test results, identify the weaknesses and vulnerabilities of the system under the intrusion of visual, sound and temperature training sets, and compare the performance differences between the system under normal conditions and when the training set is invaded. According to the analysis results, adjust and optimize the AI ​​model, and design effective defense strategies, such as improving the anti-interference ability of cameras and microphones, and improving the robustness of temperature sensors. Realize multimodal data fusion, enhance the system's comprehensive analysis capabilities of multiple sensor data, and improve the detection accuracy of the cabin environment.

[0153] Ultimately, it ensures that the internal system of the intelligent driving cockpit can operate stably under various complex and extreme conditions, and improves the detection and response capabilities of the internal cockpit environment. Through efficient, safe and comprehensive simulation testing, it reduces development costs, accelerates the R&D process, and enhances the system's defense capabilities to ensure the reliability and safety of intelligent driving technology.

[0154] Furthermore, the embodiment also includes:

[0155] Sound training set generation and circulation:

[0156] Use the sound training set generation algorithm to generate noises of different frequencies and volumes. For example, generate noises with frequencies between 200 Hz and 2000 Hz and volumes between 50 and 80 decibels as the sound training set.

[0157] These training sets of sounds are fed into a microphone system inside the cabin when simulating a scenario where a child or pet is left behind in the car.

[0158] Specific feedback responses in the simulated cockpit scenario include: the microphone system may confuse these noises with children crying or pets barking, resulting in the system being unable to accurately detect the sound signals of children or pets.

[0159] This could involve optimizing the microphone's audio recognition algorithm so that it can better distinguish between noise and real sound signals, or adding other sensors (such as infrared sensors) to help detect the presence of children or pets.

[0160] When simulating the scenario of controlling the car with commands from a voice assistant, the sound training set is used to send voice commands that can be recognized by AI but are almost imperceptible to humans (such as opening the window, turning on the air conditioner, making a call, SOS for help, etc.), which are sent to the vehicle via the broadcast radio.

[0161] Specific feedback responses in the simulated cockpit scenario include: the vehicle responds to interference commands, successfully opens the window, turns on the air conditioner, makes a phone call, etc.

[0162] Generation and circulation of training sets for interfering with the camera inside the cockpit:

[0163] Generate a training set that interferes with the camera inside the cabin, such as generating some blurred images, light and shadow interference, or patches of specific colors to disrupt the camera's recognition of children or pets.

[0164] These training sets are fed into the camera system.

[0165] Specific feedback responses within the simulated cockpit scenario include: the camera may not be able to capture the image of a child or pet clearly, or may mistakenly identify other objects as children or pets.

[0166] Measures may include optimizing the camera's image processing algorithm to improve its anti-interference ability, or using multiple cameras for multi-angle shooting to ensure that the location of children or pets can be accurately detected.

[0167] According to a second embodiment of the present invention, the present invention claims protection for a simulation test device for an intelligent driving cockpit, comprising:

[0168] one or more processors;

[0169] A memory having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the simulation test method for the intelligent driving cockpit.

[0170] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0171] In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional units. The above is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of this application, or directly or indirectly used in other related technical fields, is also included in the patent protection scope of the present application.

[0172] The specific implementation methods of the invention are described in detail above, but they are only examples, and the present application is not limited to the specific implementation methods described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of the present application, and therefore, the equalization, modification, and improvement made without departing from the spirit and principle of the present application should be included in the scope of the present application.

Claims

1. A simulation test method for an intelligent driving cockpit, characterized in that: include: Performing simulation environment configuration on the cockpit of the vehicle to be tested to obtain a training set that conforms to the cockpit; Using a camera to collect data about the interior scene of the cockpit to obtain an image of the interior scene of the cockpit; Adding the training set to the internal scene image and then inputting it into a test model, recognizing the training set to obtain a recognition result; Responding based on the recognition result, controlling the interior scene of the vehicle to be tested and verifying and optimizing the vehicle to be tested; The configuration of the simulation environment for the cockpit of the vehicle to be tested also includes: Configure the simulation environment for the cockpit of the test vehicle and use the simulation platform to create a virtual cockpit interior scene; The cockpit interior scene at least includes: Instrument panel, infotainment system, seat, door sensor, temperature sensor, live body legacy scene; Setting environmental parameters of the interior scene of the cockpit to simulate real driving conditions; The environmental parameters include at least: Light, sound; The acquiring of a training set that conforms to the cockpit further includes: Generate a training set for the cockpit internal system using a training set generation algorithm; The classification of the training set at least includes visual interference class, sound interference class, and touch screen misoperation class; Dynamically generate and apply the training set so that the test meets actual driving conditions; The method of using a camera to collect data of the interior scene of the cockpit to obtain an image of the interior scene of the cockpit also includes: When collecting data on the internal scene of the cockpit, marking the target object of the internal scene and the effective position of the target object; The step of adding the training set to the internal scene image and then inputting it into a test model, and recognizing the training set to obtain a recognition result, further includes: Adding the training set to the target object of the internal scene and the valid position of the target object to form an internal scene training image; The internal scene training image is input into the test model through the program interface, and the training set is loaded into the effective position of the target object in the preset position to realize a dynamic intrusion; The responding based on the recognition result, controlling the internal scene of the vehicle to be tested and verifying and optimizing the vehicle to be tested also includes: Apply the generated recognition results in a simulation environment to test the reaction and decision of the vehicle to be tested under the intrusion environment, and record the reaction performance of the vehicle to be tested under different intrusions; The reaction performance includes at least: Error rate, response time, system stability.

2. A simulation test method for an intelligent driving cockpit according to claim 1, characterized in that: Also includes: The target object library, training set database and driving scenario database are preset; The preset target object in the target object library is the object to be invaded; The target object library at least includes: Driver status, environment status, Face ID.

3. A simulation test method for an intelligent driving cockpit according to claim 2, characterized in that: When the cockpit interior scene is a dashboard, it also includes: After the vehicle to be tested enters the tunnel, the ambient light changes in the interior scene of the cockpit are simulated, the brightness of the dashboard of the driving cockpit is automatically adjusted, the daytime running lights are automatically turned on, and the changes in the vehicle are observed after being added to the training set.

4. The simulation test method for an intelligent driving cockpit according to claim 3, characterized in that: When the cabin interior scene is a scene left behind by a living body, it also includes: The living subject is a child or a pet; Simulating the activities of children or pets in the vehicle to be tested in different positions and states, including at least being still, moving, and making sounds; When a child is left behind, simulate the scenario of the child being left in the car, including at least moving, crying, and playing in the seat; When left behind for a pet, simulate the scenario of a pet being left in a vehicle, including at least moving, barking, and attempting to escape; Generate a training set for interfering with a camera inside a cockpit, wherein the training set is used to disrupt the camera's recognition of children or pets; Using a sound training set generation algorithm, generating a sound training set for interfering with a microphone inside the cabin, wherein the sound training set is used to interfere with the microphone's detection of a child's crying or a pet's barking; The sound training set includes noises of different frequencies and volumes to simulate interference factors in a real environment; The temperature training set generation technology is used to simulate the interference of the temperature sensor in the car, and a training set is generated to make the temperature sensor unable to accurately sense the temperature changes inside the cabin, resulting in the failure to detect the high temperature environment in time. The temperature training set is used to simulate the failure of the temperature sensor or human intervention, affecting the system's response to high temperature.

5. A simulation test device for an intelligent driving cockpit, characterized in that: Also includes: one or more processors; A memory having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement a simulation test method for an intelligent driving cockpit according to any one of claims 1 to 4.

Images