Highway high-level auxiliary driving system based on truth value system

Abstract

The invention relates to an expressway high-level auxiliary driving system based on a truth value system, and the system comprises an ADAS domain controller which comprises a sensing module and a regulation control module, and the sensing module is used for receiving the point cloud data information of a vehicle driving road, the image data information of the vehicle driving road, and the data information of the position and posture of a vehicle, and transmitting the data information to the ADAS domain controller; the sensing module is used for sensing a vehicle aided driving decision and outputting data information of a sensing result of the vehicle aided driving, and the regulation and control module is connected with the sensing module and used for receiving the data information of the sensing result of the vehicle aided driving and representing the vehicle aided driving decision by adopting a regulation and control algorithm to obtain data information of the vehicle aided driving decision. The high-level auxiliary driving cost of the whole vehicle is reduced, the use efficiency of the high-level auxiliary driving system is improved, driver fatigue caused by long-distance high-speed driving is relieved, and traffic accidents are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle auxiliary driving, in particular to a highway high-level auxiliary driving system based on a true value system. BACKGROUND

[0002] The true value system is a data acquisition system composed of millimeter wave radar, laser radar, high-precision integrated inertial navigation and other vehicle-mounted sensors plus high-efficiency data recording equipment. Since it can produce data (true value) with higher reliability than the measured sensor through data processing, the true value system is often used to evaluate the performance of the measured sensor. At the same time, based on the accuracy of data acquisition, after cleaning, labeling and data mining of true value data, a natural driving scene data set can be formed, and a natural driving scene library can be built.

[0003] With the continuous development of automatic driving technology, more and more vehicles begin to be equipped with advanced driving assistance systems (ADAS). These systems can help drivers better control vehicles in complex traffic environments, improve driving safety and comfort. However, existing ADAS systems still have some problems in actual application, such as insufficient perception accuracy, imperfect decision-making algorithms, etc. SUMMARY

[0004] In view of the above problems, the present application provides a highway high-level auxiliary driving system based on a true value system, which not only reduces the high cost of the whole vehicle high-level auxiliary driving, improves the use efficiency of the high-level auxiliary driving system, but also reduces the fatigue of the driver caused by long-distance high-speed driving and reduces the occurrence of traffic accidents.

[0005] In order to achieve the above purposes and other related purposes, the technical scheme provided by the present application is as follows: A highway high-level auxiliary driving system based on a true value system, the system comprising: a laser radar, a camera and a combined navigation, the laser radar being used for acquiring point cloud data information of a road on which a vehicle travels in real time, the camera being used for acquiring image data information of the road on which the vehicle travels in real time, and the combined navigation being used for acquiring data information of a position and an attitude of the vehicle in real time; an ADAS domain controller, comprising a perception module and a control module, the perception module being used for receiving the point cloud data information of the road on which the vehicle travels, the image data information of the road on which the vehicle travels and the data information of the position and the attitude of the vehicle, and outputting data information of a perception result of vehicle auxiliary driving, the control module being connected with the perception module and being used for receiving the data information of the perception result of vehicle auxiliary driving, representing vehicle auxiliary driving decision-making by using a control algorithm, and obtaining data information of a decision of vehicle auxiliary driving; The upper control signal interface is connected with the ADAS domain controller and is used to transmit data information of a decision of vehicle assisted driving to a vehicle end.

[0006] Further, the system further comprises a vehicle end receiving interface connected with the upper control signal interface and used to receive data information of a decision of vehicle assisted driving.

[0007] Further, the system further comprises a drive-by-wire chassis connected with the vehicle end receiving interface and used to control and dynamically adjust the chassis of the vehicle according to the data information of the decision of vehicle assisted driving.

[0008] Further, the drive-by-wire chassis communicates data with the vehicle end receiving interface through a chassis signal protocol.

[0009] Further, the adoption of a regulation and control algorithm to characterize the decision of vehicle assisted driving comprises: M1. Based on the data information of the perception result of vehicle assisted driving, a perception result data set of vehicle assisted driving is constructed and divided into a perception result data training set and a detection set of vehicle assisted driving; M2. The perception result data training set of vehicle assisted driving is input into a vehicle assisted driving decision model for training and learning to obtain a trained vehicle assisted driving decision model; M3. Based on the trained vehicle assisted driving decision model, the perception result data set of vehicle assisted driving is input to characterize the decision of vehicle assisted driving, and data information of the decision of vehicle assisted driving is obtained.

[0010] Further, in step M3, the perception result data detection set of vehicle assisted driving is input into the trained vehicle assisted driving decision model for detection, the parameters of the model are fine-tuned, and then the perception result data set of vehicle assisted driving is input to characterize the decision of vehicle assisted driving, and data information of the decision of vehicle assisted driving is obtained.

[0011] Further, the data information of the decision of vehicle assisted driving comprises data information of a desired turning angle of the vehicle, data information of acceleration of the vehicle, data information of braking torque of the vehicle, data information of lateral offset of the vehicle, and data information of longitudinal offset of the vehicle.

[0012] Further, the upper control signal interface is used for rapid mounting and dismounting, calibration, and debugging functions of high-level assisted driving, and realizes point-to-point high-level assisted driving in a highway scene after high-speed driving and before low-speed driving.

[0013] Further, the perception module is used for receiving point cloud data information of a road traveled by the vehicle, image data information of the road traveled by the vehicle, data information of a position and an attitude of the vehicle, and outputting data information of a perception result of vehicle assisted driving, which comprises the following steps: Q1. Based on the point cloud data information of the road traveled by the vehicle, the image data information of the road traveled by the vehicle, and the data information of the position and the attitude of the vehicle, noise reduction processing is performed to obtain processed point cloud data information of the road traveled by the vehicle, processed image data information of the road traveled by the vehicle, and processed data information of the position and the attitude of the vehicle; Q2. Based on the processed point cloud data information of the road traveled by the vehicle, the processed image data information of the road traveled by the vehicle, and the processed data information of the position and the attitude of the vehicle, a weighted average algorithm is used to fuse multi-sensor data of the vehicle to obtain fused data information of the multi-sensor of the vehicle; Q3. Based on the fused data information of the multi-sensor of the vehicle, a perception result model of vehicle assisted driving is constructed to characterize the perception result of vehicle assisted driving, and data information of the perception result of vehicle assisted driving is obtained.

[0014] Further, in step Q3, the fused data information of the multi-sensor of the vehicle is input into the perception result model of vehicle assisted driving for training and learning, and weights and biases of the model are optimized to obtain a trained perception result model of vehicle assisted driving.

[0015] The present application has the following positive effects: The present application adopts a sensor with appropriate precision, increases an ADAS domain controller, reserves an upper control signal interface, forms a complete high-level assisted driving system hardware, fully utilizes a true value system data acquisition and labeling function, comprehensively improves a large model data training amount of a perception module, enhances perception recognition stability, combines a regulation and control algorithm and a chassis signal protocol, achieves the purpose of controlling high-level assisted driving of the vehicle, not only reduces high-level assisted driving cost of the vehicle, improves use efficiency of the high-level assisted driving system, but also reduces driver fatigue caused by long-distance high-speed driving and reduces the occurrence of traffic accidents. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 The figure is a schematic diagram of a system framework of the present application; Figure 2 The figure is a flowchart of a regulation and control algorithm of the present application; Figure 3 The figure is a working flowchart of a perception module of the present application. DETAILED DESCRIPTION

[0017] The following description of exemplary embodiments of the present disclosure is made in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0018] Example 1: Figure 1 As shown, a high-level assisted driving system for highways based on a truth value system includes: Laser radar, camera and integrated navigation, the laser radar is used to obtain point cloud data information of the vehicle's driving road in real time, the camera is used to obtain image data information of the vehicle's driving road in real time, and the integrated navigation is used to obtain data information of the vehicle's position and posture in real time; The ADAS domain controller includes a perception module and a regulation and control module. The perception module is used to receive point cloud data information of the vehicle's driving road, image data information of the vehicle's driving road, and data information of the vehicle's position and posture, and output data information of the perception results of the vehicle's assisted driving. The regulation and control module is connected to the perception module and is used to receive data information of the perception results of the vehicle's assisted driving, characterize the vehicle's assisted driving decisions using a regulation and control algorithm, and obtain data information of the vehicle's assisted driving decisions; The upper body control signal interface is connected to the ADAS domain controller and is used to transmit data information of vehicle assisted driving decision to the vehicle end.

[0019] In this embodiment, the system further includes a vehicle-side receiving interface connected to the upper-mounted control signal interface for receiving data information for the vehicle's assisted driving decision.

[0020] In this embodiment, the system also includes a wire-controlled chassis connected to the vehicle-end receiving interface, which is used to control and dynamically adjust the vehicle's chassis based on data information of the vehicle's assisted driving decision.

[0021] In this embodiment, the wire-controlled chassis communicates and transmits data with the vehicle-end receiving interface through a chassis signal protocol.

[0022] In this embodiment, the use of a regulatory control algorithm to characterize the vehicle assisted driving decision includes: M1. Based on the data information of the perception results of the vehicle assisted driving, construct a vehicle assisted driving perception result data set, and divide it into a vehicle assisted driving perception result data training set and a test set; M2. input the perception result data set of the vehicle assisted driving into the vehicle assisted driving decision model for training and learning, to obtain a trained vehicle assisted driving decision model; M3. based on the trained vehicle assisted driving decision model, input the perception result data set of the vehicle assisted driving, to characterize the vehicle assisted driving decision, to obtain the data information of the vehicle assisted driving decision.

[0023] In this embodiment, in step M3, the perception result data set of the vehicle assisted driving is input into the trained vehicle assisted driving decision model for detection, the parameters of the model are fine-tuned, and then the perception result data set of the vehicle assisted driving is input to characterize the vehicle assisted driving decision, to obtain the data information of the vehicle assisted driving decision.

[0024] In this embodiment, the data information of the vehicle assisted driving decision includes data information of a vehicle expected turning angle, data information of a vehicle acceleration, data information of a vehicle braking torque, data information of a vehicle lateral offset, and data information of a vehicle longitudinal offset.

[0025] In this embodiment, the upper control signal interface is used for rapid loading and unloading, calibration, and debugging functions of high-level assisted driving, to realize point-to-point high-level assisted driving in the highway scene after going up to high speed and before going down to high speed.

[0026] In this embodiment, the perception module is used to receive point cloud data information of a vehicle driving road, image data information of the vehicle driving road, and data information of a vehicle position and attitude, and output data information of a vehicle assisted driving perception result, including: Q1. based on the point cloud data information of the vehicle driving road, the image data information of the vehicle driving road, and the data information of the vehicle position and attitude, perform noise reduction processing to obtain processed point cloud data information of the vehicle driving road, processed image data information of the vehicle driving road, and processed data information of the vehicle position and attitude; Q2. based on the processed point cloud data information of the vehicle driving road, the processed image data information of the vehicle driving road, and the processed data information of the vehicle position and attitude, use a weighted average algorithm to fuse the multi-sensor data of the vehicle, to obtain fused multi-sensor data information of the vehicle; Q3. based on the fused multi-sensor data information of the vehicle, construct a vehicle assisted driving perception result model, to characterize the vehicle assisted driving perception result, to obtain data information of the vehicle assisted driving perception result.

[0027] In this embodiment, in step Q3, the data information of the fused multi-sensor of the vehicle is input into the perception result model of the vehicle assisted driving for training and learning, and the weights and biases of the model are optimized to obtain the trained perception result model of the vehicle assisted driving.

[0028] Embodiment 2: Based on the highway high-level assisted driving system of the true value system in embodiment 1, the present application is further described and explained as follows.

[0029] A highway high-level assisted driving system based on a true value system, the system comprising: a laser radar, a camera and a combined navigation, the laser radar being used to acquire point cloud data information of a road on which a vehicle travels in real time, the camera being used to acquire image data information of the road on which the vehicle travels in real time, and the combined navigation being used to acquire data information of a position and a pose of the vehicle in real time; an ADAS domain controller comprising a perception module and a control module, the perception module being used to receive the point cloud data information of the road on which the vehicle travels, the image data information of the road on which the vehicle travels and the data information of the position and the pose of the vehicle, and output data information of a perception result of the vehicle assisted driving, the control module being connected with the perception module and being used to receive the data information of the perception result of the vehicle assisted driving, represent a decision of the vehicle assisted driving by using a control algorithm, and obtain data information of the decision of the vehicle assisted driving; an upper control signal interface connected with the ADAS domain controller and being used to transmit the data information of the decision of the vehicle assisted driving to a vehicle end.

[0030] In this embodiment, a true value system collects environment and motion data by using multi-sensors (laser radar, camera and combined navigation), and after the data is collected by an industrial computer, data cleaning (removing noise and redundancy) and labeling (assigning semantic information to the data) are sequentially performed, and finally the data is used for perception training to construct a precise environment perception model, to provide reliable true value data support for high-level assisted driving and other scenes, and to realize a closed-loop process from multi-source collection to intelligent training.

[0031] In this embodiment, a high-level assisted driving system is constructed to form a complete closed loop of “perception-decision-execution”. Multi-source sensors are deployed at the front end: n laser radars emit laser beams to generate three-dimensional point clouds to accurately capture the spatial structure of surrounding objects; n cameras collect visual images to provide texture and semantic information to assist in identifying traffic signs and other elements; and one combined navigation fuses GNSS and IMU to output the vehicle pose to solve the problem of positioning continuity in satellite signal blocking scenarios. The advantages of multi-sensors are complementary to meet the needs of environment and state perception.

[0032] The data is aggregated to the ADAS domain controller, is firstly fused by the perception module, is matched by an algorithm to realize feature matching, space-time alignment, constructs a unified environment cognition model, identifies obstacles, lanes, traffic participants and the like, and outputs high-precision perception results. Then, the data is combined with preset strategies such as adaptive cruise and lane keeping by the rule control module, plans a driving path, and is converted into precise control instructions such as throttle, brake and steering. Finally, the data is connected to the vehicle execution layer through a vehicle interface, drives the chassis and power system, and simultaneously feeds back the vehicle state, so as to guarantee dynamic optimization and safety redundancy of the control strategy and support cooperative operation of L2+ level advanced auxiliary driving functions.

[0033] In the embodiment, the data acquisition: the laser radar, the camera and the integrated navigation respectively collect the point cloud data, the image data and the position and attitude data around the vehicle.

[0034] Data preprocessing: the collected data is preprocessed, such as denoising, filtering and the like.

[0035] Data fusion: the preprocessed data is input into the perception module, and the perception result of the environment around the vehicle is generated through a data fusion algorithm.

[0036] 4. Decision generation: the rule control module generates the auxiliary driving decision of the vehicle according to the perception result by using a rule control algorithm.

[0037] Control signal transmission: the generated decision data is transmitted to the vehicle control system through the upper control signal interface, so as to realize the control of the vehicle.

[0038] In the embodiment, the perception module is one of the core parts of the application, and its main function is to fuse the data of multiple sensors to generate the perception result of the environment around the vehicle. The design of the perception module is as follows: Data fusion algorithm: the Kalman filter algorithm is used to fuse the data of the laser radar, the camera and the integrated navigation. The Kalman filter algorithm can effectively process multi-source data and improve the accuracy of the perception result.

[0039] Target detection: the deep learning algorithm is used to detect the target in the image data collected by the camera, and the vehicle, the pedestrian, the obstacle and the like are identified.

[0040] Lane line detection: the lane line detection algorithm based on the Sobel operator and the Roberts operator is used to extract the lane line information from the image data.

[0041] The application further provides a computer readable storage medium, and the computer readable storage medium stores a computer program programmed or configured to execute any one of the high-speed highway advanced auxiliary driving methods based on the true value system.

[0042] Any reference to storage, memory, database or other medium herein can include non-volatile and / or volatile storage. Non-volatile storage can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile storage can include random-access memory (RAM), or external cache memory. By way of illustration, and not limitation, RAM is available in many forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct Rambus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM). The disclosure should make it manifestly clear that the scope of the disclosure is made not subject to the RAM types recited herein.

[0043] In summary, the application not only reduces the cost of high-level auxiliary driving of the whole vehicle and improves the use efficiency of the high-level auxiliary driving system, but also reduces the fatigue of the driver caused by long-distance high-speed driving and reduces the occurrence of traffic accidents.

[0044] The foregoing detailed description has not been presented to limit the scope of the present disclosure. Various modifications and changes can be made to the embodiments described without departing from the spirit and principles of the disclosure. Any modifications, equivalent substitutions, improvements, and the like made within the spirit and principles of the present disclosure should be included in the scope of the present disclosure.

Claims

1. A high-level highway driver assistance system based on a truth value system, characterized in that: The system comprises: Laser radar, camera and integrated navigation, the laser radar is used to obtain point cloud data information of the vehicle's driving road in real time, the camera is used to obtain image data information of the vehicle's driving road in real time, and the integrated navigation is used to obtain data information of the vehicle's position and posture in real time; The ADAS domain controller includes a perception module and a regulation and control module. The perception module is used to receive point cloud data information of the vehicle's driving road, image data information of the vehicle's driving road, and data information of the vehicle's position and posture, and output data information of the perception results of the vehicle's assisted driving. The regulation and control module is connected to the perception module and is used to receive data information of the perception results of the vehicle's assisted driving, characterize the vehicle's assisted driving decisions using a regulation and control algorithm, and obtain data information of the vehicle's assisted driving decisions; The body control signal interface is connected to the ADAS domain controller and is used to transmit data information of vehicle assisted driving decision to the vehicle end; The characterization of the vehicle assisted driving decision by using the regulation and control algorithm includes: M1. Based on the data information of the perception results of the vehicle assisted driving, construct a vehicle assisted driving perception result data set, and divide it into a vehicle assisted driving perception result data training set and a test set; M2. Input the vehicle-assisted driving perception result data training set into the vehicle-assisted driving decision model for training and learning to obtain a trained vehicle-assisted driving decision model; M3. Based on the trained vehicle-assisted driving decision model, a vehicle-assisted driving perception result dataset is input, the vehicle-assisted driving decision is characterized, and data information of the vehicle-assisted driving decision is obtained.

2. The highway high-level driver assistance system based on the truth value system according to claim 1 is characterized in that: The system also includes a vehicle-side receiving interface connected to the upper-mounted control signal interface for receiving data information for the vehicle's assisted driving decision-making.

3. The highway high-level driver assistance system based on the truth value system according to claim 2 is characterized in that: The system also includes a wire-controlled chassis connected to the vehicle-side receiving interface, which is used to control and dynamically adjust the vehicle's chassis based on data information of the vehicle's assisted driving decision.

4. The highway high-level driver assistance system based on the truth value system according to claim 3 is characterized by: The wire-controlled chassis communicates and transmits data with the vehicle-end receiving interface through a chassis signal protocol.

5. The highway high-level driver assistance system based on the truth value system according to claim 1 is characterized in that: In step M3, the perception result data detection set of the vehicle assisted driving is input into the trained vehicle assisted driving decision model for detection, the parameters of the model are fine-tuned, and then the perception result data set of the vehicle assisted driving is input to characterize the vehicle assisted driving decision and obtain the data information of the vehicle assisted driving decision.

6. The highway high-level driver assistance system based on the truth value system according to claim 1 is characterized by: The data information for the vehicle assisted driving decision includes data information of the vehicle's expected turning angle, data information of the vehicle's acceleration, data information of the vehicle's braking torque, data information of the vehicle's lateral offset, and data information of the vehicle's longitudinal offset.

7. The highway high-level driver assistance system based on the truth value system according to claim 1 is characterized by: The upper installation control signal interface is used for the rapid loading and unloading, calibration, and debugging functions of high-level assisted driving, realizing point-to-point high-level assisted driving in highway scenarios after getting on and before getting off the highway.

8. The highway high-level driver assistance system based on the truth value system according to claim 1 is characterized in that: The perception module is used to receive point cloud data information of the vehicle's driving road, image data information of the vehicle's driving road, and data information of the vehicle's position and posture, and outputs data information of the perception results of the vehicle assisted driving, including: Q1. Performing noise reduction processing based on the point cloud data information of the vehicle's travel path, the image data information of the vehicle's travel path, and the data information of the vehicle's position and posture to obtain processed point cloud data information of the vehicle's travel path, the image data information of the vehicle's travel path, and the data information of the vehicle's position and posture; Q2. Based on the processed point cloud data information of the vehicle's travel path, the image data information of the vehicle's travel path, and the data information of the vehicle's position and posture, a weighted averaging algorithm is used to fuse the vehicle's multi-sensor data to obtain fused vehicle multi-sensor data information; Q3. Based on the fused data information from multiple sensors of the vehicle, a perception result model of vehicle-assisted driving is constructed, the perception result of vehicle-assisted driving is characterized, and data information of the perception result of vehicle-assisted driving is obtained.

9. The highway high-level driver assistance system based on the truth value system according to claim 8 is characterized in that: In step Q3, the fused data information of the vehicle's multiple sensors is input into the vehicle-assisted driving perception result model for training and learning, and the weights and biases of the model are optimized to obtain a trained vehicle-assisted driving perception result model.

Images