Automatic driving method and system
By using roadside units to generate track models and path planning in remote control racing, using deep learning networks and signal receiving devices, the problems of high device complexity and cost in the prior art are solved, autonomous driving is realized and racing is promoted diversified development.
Patent Information
- Application Number
- CN202210775183.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-07-01
- Publication Date
- 2024-12-10
- Estimated Expiration
- 2042-07-01
AI Technical Summary
The existing remote-controlled racing cars have the complex structure and high cost of autonomous driving, which limits the diversified development of racing cars.
The track model generation and path planning are used to use roadside units to enable the vehicle to realize autonomous driving in the track through signal receiving devices, reducing the complexity and cost of the racing side equipment, and using deep learning network to build the track model, combining vehicle positioning and Control the characteristic information.
It realizes the autonomous driving of remote-controlled racing cars, reduces the complexity and cost of the equipment, increases the technological and fun of the racing cars, and promotes the diversified development of racing cars.
Smart Images
Figure CN115167416B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of vehicle control technology, and more specifically, relates to an automatic driving method and system. Background Art
[0002] As the research on autonomous driving continues to gain popularity, the concept of autonomous driving has spread to different fields, such as indoor wheeled robots, driverless formula racing cars, etc. At present, the research on autonomous driving is mainly focused on single-vehicle autonomous driving, and trajectory tracking corresponds to single-vehicle intelligence, where the single vehicle itself performs path planning and tracking. Existing remote control racing cars mostly use single-vehicle autonomous driving to achieve autonomous driving, which results in a more complex structure and higher cost for the remote control racing device, which is not conducive to promoting the diversified development of racing cars. Summary of the invention
[0003] The present invention provides an automatic driving method and system to solve the problem that the device structure of the automatic driving of a remote-controlled racing car in the prior art is relatively complex and the cost is relatively high.
[0004] The present invention provides an automatic driving method, comprising:
[0005] The track model generation module of the roadside unit constructs a track model according to the track image information; the path planning module of the roadside unit performs global path planning according to the track model, obtains the global planned driving path, and sends it to the path tracking module of the roadside unit;
[0006] The positioning tag of the vehicle side unit sends the vehicle positioning information to the positioning base station of the road side unit, and the positioning base station receives the vehicle positioning information and sends it to the path tracking module;
[0007] The vehicle characteristic detection module of the vehicle side unit obtains vehicle characteristic information, and the second signal transceiver module of the vehicle side unit sends the vehicle characteristic information to the first signal transceiver module of the road side unit; the first signal transceiver module receives the vehicle characteristic information and sends it to the path tracking module;
[0008] The path tracking module obtains vehicle control information according to the global planned driving path, the vehicle positioning information and the vehicle characteristic information, and sends the vehicle control information to the first signal transceiver module;
[0009] The first signal transceiver module sends the vehicle control information to the second signal transceiver module; the second signal transceiver module receives the vehicle control information and sends it to the driving control module of the vehicle side unit; the driving control module controls the driving of the vehicle body based on the vehicle control information.
[0010] Preferably, the vehicle characteristic detection module obtains the heading information of the vehicle body through an IMU heading detection device, and the vehicle characteristic detection module obtains the speed information and acceleration information of the vehicle body through a speed and acceleration detection module.
[0011] Preferably, the driving control module is connected to a motor controller and a steering gear controller on the vehicle body respectively, and the driving control module controls the motor controller and the steering gear controller based on the vehicle control information.
[0012] Preferably, the track model generation module constructs the track model based on a trained deep learning network.
[0013] Preferably, the trained deep learning network is obtained by: constructing a data set based on a number of original images of the track and their corresponding track label images; according to the data set, optimizing the parameters in the deep learning network based on the loss function to obtain the trained deep learning network.
[0014] On the other hand, the present invention provides an automatic driving system, comprising: a roadside unit and a vehicle side unit; the roadside unit is arranged beside a track, and the vehicle side unit is located in the track; the roadside unit comprises a track model generation module, a path planning module, a positioning base station, a path tracking module and a first signal transceiver module; the vehicle side unit comprises a vehicle body, and a second signal transceiver module, a positioning tag, a vehicle characteristic detection module and a driving control module installed on the vehicle body;
[0015] The track model generation module is used to construct a track model based on the track image information; the path planning module is used to perform global path planning based on the track model to obtain a globally planned driving path; the positioning base station is used to receive vehicle positioning information; the path tracking module is used to obtain vehicle control information based on the globally planned driving path, vehicle positioning information and vehicle characteristic information; the first signal transceiver module is used to send vehicle control information to the second signal transceiver module; the second signal transceiver module is used to send vehicle characteristic information to the first signal transceiver module; the positioning tag is used to send vehicle positioning information; the vehicle characteristic detection module is used to obtain vehicle characteristic information; the driving control module is used to control the driving of the vehicle body based on the vehicle control information.
[0016] Preferably, the vehicle characteristic detection module includes: an IMU heading detection device and a speed and acceleration detection module; the IMU heading detection device is used to obtain the vehicle's heading information, and the speed and acceleration detection module is used to obtain the vehicle's speed information and the vehicle's acceleration information.
[0017] Preferably, the driving control module is connected to a motor controller and a steering gear controller on the vehicle body.
[0018] Preferably, the roadside unit also includes: an image acquisition module; the image acquisition module is used to acquire track image information and send it to the track model generation module; the track model generation module has a built-in trained deep learning network, and the track model generation module constructs a track model based on the acquired track image information using the trained deep learning network.
[0019] One or more technical solutions provided in the present invention have at least the following technical effects or advantages:
[0020] The roadside unit arranged beside the track of the present invention includes a track model generation module, a path planning module, a positioning base station, a path tracking module and a first signal transceiver module, and the vehicle side unit located in the track includes a vehicle body and a second signal transceiver module installed on the vehicle body, a positioning tag, a vehicle characteristic detection module and a driving control module. The present invention performs path planning and trajectory tracking through the roadside unit, and the vehicle travels based on the vehicle control information from the roadside unit, which can reduce the complexity of the device for realizing automatic driving of the remote control racing car and reduce the cost. That is, the present invention proposes a method for automatic driving of a racing car without perception based on the roadside unit, which incorporates the idea of the Internet of Vehicles, performs path planning and trajectory tracking calculations through the roadside unit, so that the racing car itself does not need to be installed with complex devices and calculation modules, and can realize automatic driving in the track mainly through the signal receiving device, which not only reduces the complexity of the racing car side device and reduces the cost, but also greatly increases the scientific and interesting nature of the racing car, and can promote the diversified development of remote control racing cars. BRIEF DESCRIPTION OF THE DRAWINGS
[0021] Figure 1 A schematic diagram of a framework of an automatic driving system provided in Embodiment 1 of the present invention;
[0022] Figure 2 A schematic diagram of the positions of a roadside unit, a vehicle side unit and a track in an automatic driving system provided in Example 1 of the present invention. DETAILED DESCRIPTION
[0023] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.
[0024] Embodiment 1:
[0025] Embodiment 1 provides an automatic driving system, see Figure 1 , Figure 2, including: a roadside unit 1 and a vehicle side unit 2; the roadside unit 1 is arranged beside the track, and the vehicle side unit 2 is located in the track. The roadside unit 1 includes a track model generation module 101, a path planning module 102, a path tracking module 103, a positioning base station 104 and a first signal transceiver module 105. The vehicle side unit 2 includes a vehicle body, and a second signal transceiver module 201, a positioning tag 202, a vehicle characteristic detection module 203 and a driving control module 204 installed on the vehicle body.
[0026] The track model generation module 101 is used to construct a track model according to the track image information. The path planning module 102 is used to perform global path planning according to the track model to obtain a globally planned driving path. The path tracking module 103 is used to obtain vehicle control information according to the globally planned driving path, vehicle positioning information and vehicle characteristic information. The positioning base station 104 is used to receive vehicle positioning information. The first signal transceiver module 105 is used to send vehicle control information to the second signal transceiver module 201. The second signal transceiver module 201 is used to send vehicle characteristic information to the first signal transceiver module 105. The positioning tag 202 is used to send vehicle positioning information. The vehicle characteristic detection module 203 is used to obtain vehicle characteristic information. The driving control module 204 is used to control the driving of the vehicle body based on the vehicle control information.
[0027] In addition, the roadside unit 1 may further include: an image acquisition module; the image acquisition module is used to acquire track image information and send it to the track model generation module 101. The track model generation module 101 has a built-in trained deep learning network, and the track model generation module 101 uses the trained deep learning network to build a track model based on the acquired track image information. For example, the image acquisition module uses a camera, which is connected to the track model generation module 101 and is used to take pictures of the track to input into the deep learning network.
[0028] The vehicle characteristic detection module 203 includes: an IMU heading detection device and a speed and acceleration detection module; the IMU heading detection device is used to obtain the vehicle heading information, and the speed and acceleration detection module is used to obtain the vehicle speed information and the vehicle acceleration information.
[0029] The driving control module 204 is connected to the motor controller and the steering gear controller on the vehicle body.
[0030] The present invention performs path planning and trajectory tracking through a roadside unit, and the racing car can travel based on the vehicle control information from the roadside unit. The design of the above system can reduce the complexity of the device for realizing automatic driving of the remote-controlled racing car and reduce the cost.
[0031] Embodiment 2:
[0032] Example 2 is an automatic driving method implemented using the automatic driving system provided by Example 1.
[0033] Specifically, embodiment 2 provides an automatic driving method, including:
[0034] The track model generation module of the roadside unit constructs a track model according to the track image information; the path planning module of the roadside unit performs global path planning according to the track model, obtains the global planned driving path, and sends it to the path tracking module of the roadside unit;
[0035] The positioning tag of the vehicle side unit sends the vehicle positioning information to the positioning base station of the road side unit, and the positioning base station receives the vehicle positioning information and sends it to the path tracking module;
[0036] The vehicle characteristic detection module of the vehicle side unit obtains vehicle characteristic information, and the second signal transceiver module of the vehicle side unit sends the vehicle characteristic information to the first signal transceiver module of the road side unit; the first signal transceiver module receives the vehicle characteristic information and sends it to the path tracking module;
[0037] The path tracking module obtains vehicle control information according to the global planned driving path, the vehicle positioning information and the vehicle characteristic information, and sends the vehicle control information to the first signal transceiver module;
[0038] The first signal transceiver module sends the vehicle control information to the second signal transceiver module; the second signal transceiver module receives the vehicle control information and sends it to the driving control module of the vehicle side unit; the driving control module controls the driving of the vehicle body based on the vehicle control information.
[0039] Among them, the vehicle characteristic detection module obtains the heading information of the vehicle body through the IMU heading detection device, and the vehicle characteristic detection module obtains the speed information and acceleration information of the vehicle body through the speed and acceleration detection module.
[0040] The driving control module is connected to the motor controller and the steering gear controller on the vehicle body respectively, and the driving control module controls the motor controller and the steering gear controller based on the vehicle control information.
[0041] The track model generation module constructs a track model based on a trained deep learning network.
[0042] The trained deep learning network can be obtained in the following manner: construct a data set based on a number of original images of the track and their corresponding track label images; according to the data set, optimize the parameters in the deep learning network based on the loss function to obtain the trained deep learning network.
[0043] The present invention is further described below.
[0044] The present invention uses a large number of track images to train a deep learning network for detecting and identifying the track and generating a track model. The trained deep learning network can extract track features and identify track edges. When a new track image is input, the trained deep learning network can output a track model in which feasible areas and track edges are marked. The path planning module plans a global path based on the obtained track model using a global path planning algorithm. The position of the car is detected by the positioning device in the vehicle side unit, and the speed and acceleration of the car are detected by the speed and acceleration detection module in the vehicle side unit, and then both are transmitted to the road side unit. The path tracking module in the road side unit calculates the control instructions based on the obtained global planned path, car position information, heading information, speed and acceleration information, and sends them to the vehicle side unit through a signal transceiver to control the car to track the global path.
[0045] The number of roadside units in the present invention can be determined according to the size of the track, and the detection range of the positioning device in the roadside unit needs to cover the track. The signal transceiver module in the present invention is used to realize data exchange between the roadside unit and the vehicle side unit. The positioning module in the present invention needs to use a high-precision positioning device, such as RTK, UWB, etc.
[0046] The vehicle body in the present invention can be specifically a remote-controlled racing car, which has a conventional wire-controlled chassis and is additionally equipped with a second signal transceiver module, a positioning tag, an IMU heading detection device, a speed and acceleration detection module, and a driving control module. The second signal transceiver module communicates with the first signal transceiver module in the roadside unit. The positioning tag is paired with a high-precision positioning device (such as a positioning base station) in the roadside unit. The IMU heading detection device is rigidly mounted on the racing car body to detect the heading angle of the racing car.
[0047] The present invention can achieve automatic driving by modifying the existing racing car, which can reduce the complexity of the racing car side device, reduce costs, and increase the playability, fun and technology of the racing car, which can promote the diversified development of remote control racing cars. In addition, the present invention is applicable to cars of different models, and only requires matching signal transceiver modules, so the requirements for automatic driving are low and easy to implement. The present invention can be applied to different tracks, and the track modeling can be performed using a trained deep learning model, which is easy to use and has many application scenarios.
[0048] Finally, it should be noted that the above specific implementation methods are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.
Claims
1. An automatic driving method, characterized in that: include: The track model generation module of the roadside unit constructs a track model based on the track image information; The path planning module of the roadside unit performs global path planning according to the track model, obtains a global planned driving path, and sends it to the path tracking module of the roadside unit; The positioning tag of the vehicle side unit sends the vehicle positioning information to the positioning base station of the road side unit, and the positioning base station receives the vehicle positioning information and sends it to the path tracking module; The vehicle characteristic detection module of the vehicle side unit obtains vehicle characteristic information, and the second signal transceiver module of the vehicle side unit sends the vehicle characteristic information to the first signal transceiver module of the road side unit; the first signal transceiver module receives the vehicle characteristic information and sends it to the path tracking module; The path tracking module obtains vehicle control information according to the global planned driving path, the vehicle positioning information and the vehicle characteristic information, and sends the vehicle control information to the first signal transceiver module; The first signal transceiver module sends the vehicle control information to the second signal transceiver module; the second signal transceiver module receives the vehicle control information and sends it to the driving control module of the vehicle side unit; The driving control module controls the driving of the vehicle body based on the vehicle control information.
2. The automatic driving method according to claim 1, characterized in that: The vehicle characteristic detection module obtains the heading information of the vehicle body through the IMU heading detection device, and the vehicle characteristic detection module obtains the speed information and acceleration information of the vehicle body through the speed and acceleration detection module.
3. The automatic driving method according to claim 1, characterized in that: The driving control module is connected to the motor controller and the steering gear controller on the vehicle body respectively, and the driving control module controls the motor controller and the steering gear controller based on the vehicle control information.
4. The automatic driving method according to claim 1, characterized in that: The track model generation module constructs a track model based on a trained deep learning network.
5. The automatic driving method according to claim 4, characterized in that: The trained deep learning network is obtained by: constructing a data set based on a number of original images of the track and their corresponding track label images; according to the data set, optimizing the parameters in the deep learning network based on the loss function to obtain the trained deep learning network.
6. An automatic driving system, characterized in that: include: A roadside unit and a vehicle side unit; the roadside unit is arranged beside the track, and the vehicle side unit is located inside the track; the roadside unit comprises a track model generation module, a path planning module, a positioning base station, a path tracking module and a first signal transceiver module; the vehicle side unit comprises a vehicle body, and a second signal transceiver module, a positioning tag, a vehicle characteristic detection module and a driving control module installed on the vehicle body; The track model generation module is used to construct a track model according to the track image information; The path planning module is used to perform global path planning according to the track model to obtain a global planned driving path; The positioning base station is used to receive vehicle positioning information; The path tracking module is used to obtain vehicle control information based on the global planned driving path, vehicle positioning information and vehicle characteristic information; The first signal transceiver module is used to send vehicle control information to the second signal transceiver module; The second signal transceiver module is used to send vehicle characteristic information to the first signal transceiver module; The positioning tag is used to send vehicle positioning information; The vehicle characteristic detection module is used to obtain vehicle characteristic information; The driving control module is used to control the driving of the vehicle body based on vehicle control information.
7. The automatic driving system according to claim 6, characterized in that: The vehicle characteristic detection module includes: an IMU heading detection device and a speed and acceleration detection module; the IMU heading detection device is used to obtain the vehicle heading information, and the speed and acceleration detection module is used to obtain the vehicle speed information and the vehicle acceleration information.
8. The automatic driving system according to claim 6, characterized in that: The driving control module is connected to a motor controller and a steering gear controller on the vehicle body.
9. The automatic driving system according to claim 6, characterized in that: The roadside unit also includes: an image acquisition module; the image acquisition module is used to acquire track image information and send it to the track model generation module; the track model generation module has a built-in trained deep learning network, and the track model generation module constructs a track model based on the acquired track image information using the trained deep learning network.