CN-121973806-A - Intelligent driving system fault control method, device and equipment based on generation type AI and storage medium

CN121973806ACN 121973806 ACN121973806 ACN 121973806ACN-121973806-A

Abstract

The application discloses a method, a device, equipment and a storage medium for controlling intelligent driving system faults based on a generated AI, which relate to the technical field of intelligent network-connected automobile safety and comprise the following steps: acquiring fault data and current scene data of an intelligent driving system when a vehicle runs, wherein the fault data comprises a fault type, a fault component and a signal fluctuation parameter, and the current scene data comprises an environment scene parameter, a vehicle state parameter and position information; performing fault risk association matching under different scenes according to the fault data and the current scene data to obtain corresponding fault risk types, performing matching analysis on the fault risk types according to a preset generation type AI model to obtain a vehicle component control instruction and a risk early warning instruction, and performing fault control on the intelligent driving system according to the vehicle component control instruction and the risk early warning instruction. Through the steps, the risk prevention and control predictability, accuracy and safety are improved.

Inventors

ZHANG QIYU
WANG ZHE
PAN TAO
GAO ZHONGFANG
XIAO YANG

Assignees

上汽通用五菱汽车股份有限公司

Dates

Publication Date: 20260505
Application Date: 20260109

Claims (10)

1. The intelligent driving system fault control method based on the generation type AI is characterized by comprising the following steps: Acquiring fault data and current scene data of an intelligent driving system when a vehicle runs, wherein the fault data comprises a fault type, a fault component and a signal fluctuation parameter, and the current scene data comprises an environment scene parameter, a vehicle state parameter and position information; performing fault risk association matching under different scenes according to the fault data and the current scene data to obtain corresponding fault risk types; performing matching analysis on the fault risk type according to a preset generation type AI model to obtain a vehicle component control instruction and a risk early warning instruction; and performing fault control on the intelligent driving system according to the vehicle component control instruction and the risk early warning instruction.
2. The method of claim 1, wherein the step of obtaining fault data and current scenario data for the intelligent drive system while the vehicle is in operation comprises: acquiring sensor data and controller data when a vehicle runs; Carrying out signal fluctuation detection on the sensor data and the controller data to obtain abnormal fluctuation data, signal loss data and error percentages corresponding to different sensors and controllers; determining a fault type, a fault component and a signal fluctuation parameter according to the abnormal fluctuation data, the signal loss data and the error percentage; Extracting environmental scene parameters including current illumination intensity, backlight angle, color temperature deviation, obstacle reflectivity, obstacle distance and electromagnetic interference intensity according to the sensor data; and extracting position information and vehicle state parameters including current vehicle speed, steering wheel rotation angle, gear and yaw rate according to the controller data.
3. The method of claim 1, wherein the step of performing fault risk association matching in different scenarios according to the fault data and the current scenario data to obtain the corresponding fault risk type comprises: Matching the fault data and the current scene data with a pre-constructed fault and risk scene association knowledge base to obtain matching similarity; determining a target fault type according to the matching similarity and a preset matching similarity threshold; and matching the current scene data item by item with the environment scene parameters, the vehicle state parameters and the risk threshold intervals of the position information corresponding to the target fault type to obtain a fault risk type.
4. The method of claim 3, wherein the step of matching the current scenario data item by item with the environmental scenario parameter, the vehicle state parameter, and the risk threshold interval of the location information corresponding to the target fault type to obtain a fault risk type comprises: Determining that the fault risk type is the target fault type when the current scene data is in a risk threshold interval of the environment scene parameter, the vehicle state parameter and the position information corresponding to the target fault type; When the current scene data is not in the risk threshold interval of the environment scene parameter, the vehicle state parameter and the position information corresponding to the target fault type, marking the combination of the current scene data and the target fault type as a pending type; and carrying out expansion fault updating on the undetermined type according to a preset federal learning strategy to obtain a fault risk type.
5. The method of claim 1, wherein the step of performing a matching analysis on the fault risk type according to a preset generation AI model to obtain a vehicle component control command and a risk early warning command comprises: Inputting the fault risk type, the fault data and the current scene data into a preset generation type AI model to obtain a parameter control strategy corresponding to the fault risk type; Carrying out vehicle level analysis according to the parameter control strategy to obtain a sensor layer gain adjustment value, an actuator layer target control quantity and a decision layer path correction quantity; Generating a vehicle component control command according to the sensor layer gain adjustment value, the actuator layer target control amount and the decision layer path correction amount; And generating a risk early warning instruction according to the fault risk type and a preset risk level mapping table.
6. The method of claim 5, wherein the step of generating risk early warning instructions from the fault risk type and a preset risk level map comprises: determining a risk level color code, a response time limit, a voice broadcast text and an alarm frequency corresponding to the fault risk type according to the fault risk type and a preset risk level mapping table; Generating an image-text early warning template which is matched with a vehicle-mounted display area according to the risk grade color code, and writing the response time limit into the image-text early warning template to obtain visual early warning information; and packaging the visual early warning information, the voice broadcast text and the warning frequency into a risk early warning instruction.
7. The method of claim 1, wherein the step of fault controlling the intelligent drive system in accordance with the vehicle component control instructions and the risk early warning instructions comprises: Controlling corresponding sensors, actuators and domain controllers according to the vehicle component control instructions so that the sensors adjust sensor gains according to sensor layer gain adjustment values in the vehicle component control instructions, the actuators correct actuator outputs according to actuator layer target control amounts in the vehicle component control instructions, and the domain controllers update path planning offset according to decision layer path correction amounts in the vehicle component control instructions; And controlling the corresponding man-machine interaction unit to perform fault display and/or voice early warning according to the risk early warning instruction, wherein the man-machine interaction unit comprises at least one of a vehicle-mounted display, a central control display screen and a voice broadcasting module.
8. An intelligent driving system fault control device based on a generation AI, which is characterized by comprising: The system comprises a data acquisition module, a control module and a control module, wherein the data acquisition module is used for acquiring fault data of an intelligent driving system and current scene data when a vehicle runs, the fault data comprise fault types, fault components and signal fluctuation parameters, and the current scene data comprise environment scene parameters, vehicle state parameters and position information; The fault analysis module is used for carrying out fault risk association matching under different scenes according to the fault data and the current scene data to obtain corresponding fault risk types; The risk analysis module is used for carrying out matching analysis on the fault risk type according to a preset generation type AI model to obtain a vehicle component control instruction and a risk early warning instruction; and the fault control module is used for carrying out fault control on the intelligent driving system according to the vehicle component control instruction and the risk early warning instruction.
9. A generated AI-based intelligent drive system fault control device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program configured to implement the steps of the generated AI-based intelligent drive system fault control method of any of claims 1-7.
10. A storage medium, characterized in that the storage medium is a computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the steps of the intelligent driving system failure control method based on the generated AI according to any one of claims 1 to 7.

Description

Intelligent driving system fault control method, device and equipment based on generation type AI and storage medium Technical Field The application relates to the technical field of intelligent network-connected automobile safety, in particular to an intelligent driving system fault control method, device and equipment based on a generated AI and a storage medium. Background With the development of intelligent driving technology to a higher level (L2-L4), the complexity of system software and hardware is greatly improved, and the failure mode is changed from single hardware failure to multi-component coupling deviation. The main current fault processing mode in the current industry generally stays in the post-response stage of triggering protection after the fault occurs, effective early warning and intervention are difficult to perform before the fault occurs, and an active prevention and control system with the linkage of the fault occurrence precursors and the risk scene is not constructed. The prior art mainly has the following defects that firstly, the prior risk pre-judging capability is lost. The existing scheme (such as a technology for triggering control response based on fault codes) can only respond to the occurred faults, cannot establish a mapping relation between fault characteristic parameters (such as sensor data fluctuation) and potential risk scenes (such as severe weather and environments), and cannot identify and pre-warn hidden dangers that the faults are not triggered but the risks exist. Second, risk assessment lacks a scenerization quantification. The existing risk assessment model fuses environment and vehicle behavior data, but specific fault types are not used as core variables and are associated with exclusive scene parameters, so that risk assessment results are general, and early warning instructions lack scene suitability. Finally, the safety prompts are not personalized and the operation floor is poor. Most of the safety prompts provided by the existing system are standardized texts, and parameterized and executable operation guidelines cannot be generated by combining real-time and specific scene parameters, so that users are difficult to take effective countermeasures according to the safety prompts. Therefore, how to improve predictability, accuracy and safety of risk prevention and control of intelligent driving systems is a problem to be solved. Disclosure of Invention The application mainly aims to provide a method, a device, equipment and a storage medium for controlling intelligent driving system faults based on a generated AI, and aims to solve the technical problems of improving predictability, accuracy and safety of intelligent driving system risk prevention and control. In order to achieve the above purpose, the present application provides a method for controlling a fault of an intelligent driving system based on a generated AI, the method comprising: Acquiring fault data and current scene data of an intelligent driving system when a vehicle runs, wherein the fault data comprises a fault type, a fault component and a signal fluctuation parameter, and the current scene data comprises an environment scene parameter, a vehicle state parameter and position information; performing fault risk association matching under different scenes according to the fault data and the current scene data to obtain corresponding fault risk types; performing matching analysis on the fault risk type according to a preset generation type AI model to obtain a vehicle component control instruction and a risk early warning instruction; and performing fault control on the intelligent driving system according to the vehicle component control instruction and the risk early warning instruction. In one embodiment, the step of obtaining fault data and current scene data of the intelligent driving system when the vehicle is running includes: acquiring sensor data and controller data when a vehicle runs; Carrying out signal fluctuation detection on the sensor data and the controller data to obtain abnormal fluctuation data, signal loss data and error percentages corresponding to different sensors and controllers; determining a fault type, a fault component and a signal fluctuation parameter according to the abnormal fluctuation data, the signal loss data and the error percentage; Extracting environmental scene parameters including current illumination intensity, backlight angle, color temperature deviation, obstacle reflectivity, obstacle distance and electromagnetic interference intensity according to the sensor data; and extracting position information and vehicle state parameters including current vehicle speed, steering wheel rotation angle, gear and yaw rate according to the controller data. In an embodiment, the step of performing fault risk association matching under different scenes according to the fault data and the current scene data to obtain a corresponding fault risk type includes: M