Automatic driving vehicle control method suitable for underground parking lot

Abstract

The invention discloses an automatic driving vehicle control method suitable for an underground parking lot, relates to the technical field of automatic driving, and aims to simultaneously meet the requirements of a lane driving scene with a small posture change amplitude and a parking scene in which the posture needs to be greatly adjusted in a narrow space. The obstacle avoidance control method and the vehicle parking control method both comprise a vehicle speed control system. The obstacle avoidance control method comprises the following steps that A1, an underground parking mode is started; a2, a roof high-position camera is lifted, a whole vehicle radar is started, and the vehicle speed is reduced to 5 KM / H; a3, the camera scans the surrounding environment, and the radar scans obstacles and moving objects; a4, scanning a heat source, scanning a human face and scanning the shape of an obstacle; a5, the moving speed of the moving object is scanned, and the volumes of the moving object and the obstacle are scanned; and A6, the moving obstacles and the non-moving obstacles are avoided.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to an autonomous driving vehicle control method suitable for an underground parking lot. Background Art

[0002] Autonomous driving refers to the technology that enables the car to perceive the environment, plan its path and autonomously control the vehicle.

[0003] Commonly used control algorithms include PID (Proportional-Integral-Derivative) control algorithm, LQR (linear quadratic regulator) control algorithm and MPC (Model Predictive Control) control algorithm.

[0004] 1) PID control algorithm

[0005] PID control algorithm is a classic control method widely used in automatic control systems. It is used to adjust the output of the system so that it can stably reach the expected value. PID is a combination of three control items: proportional, integral, and derivative. Each control item corresponds to a parameter, namely proportional gain (K_p), integral time (K_i), and derivative time (K_d).

[0006] 2) LQR control algorithm

[0007] LQR is an optimization control algorithm for linear systems, which aims to minimize system performance indicators by weighing system state and control input. The LQR control algorithm theoretically provides an optimal control solution for linear, time-invariant systems.

[0008] 3) MPC control algorithm

[0009] MPC is an advanced control strategy that optimizes control inputs to achieve the minimum value of a certain performance indicator by predicting the future behavior of the system in multiple steps. MPC is a control method suitable for nonlinear and time-varying systems and is often used in complex industrial and automation systems.

[0010] At present, PID control algorithm, LQR control algorithm and MPC control algorithm are widely used in the city navigation assisted driving function and automatic parking function of autonomous vehicles. In the city navigation assisted driving function scenario, the vehicle always keeps moving forward, and the vehicle's posture changes during driving are small; in the automatic parking function scenario, the vehicle needs to be able to drive in reverse gear, and the vehicle needs to make large posture adjustments in a small space. Based on the above situation, the three existing algorithms currently have the following two major defects:

[0011] High software deployment cost:

[0012] For these two different functional scenarios, the above three algorithms need to be adapted and debugged with targeted parameters. Different algorithms or different parameters need to be applied in different functional scenarios, that is, the vehicle controller needs to switch the control algorithm according to different scenarios. This results in the need to develop two sets of algorithms and adapt two sets of parameters, which increases the software deployment cost of the assisted driving system.

[0013] The user experience of autonomous parking function in underground parking lots is poor:

[0014] In the autonomous valet parking function scenario of an underground parking lot, the vehicle will automatically cruise in the lane, and when the vehicle reaches the vicinity of the parking space, the vehicle can automatically park in the parking space. That is, the autonomous valet parking function of the underground parking lot has both lane driving scenarios with small posture changes and parking scenarios that require large posture adjustments in a small space. Using the current control algorithm will cause the vehicle controller to switch the control algorithm according to different scenarios. When switching, the vehicle needs to stop completely before switching to another control algorithm, which reduces parking efficiency. At the same time, different control algorithms have different comfort experiences, resulting in a poor user experience.

[0015] To sum up, the conventional algorithms currently used in autonomous driving cannot simultaneously meet the requirements of lane driving scenarios with small posture changes and parking scenarios that require large posture adjustments in a small space. In addition, there are currently many intersections in underground parking lots, and pedestrians appear irregularly, there are no dedicated sidewalks, and automatic assisted driving is difficult to identify. Autonomous driving has huge safety hazards and weak ability to find parking spaces. Therefore, the market urgently needs to develop an autonomous driving vehicle control method suitable for underground parking lots to help people solve existing problems. Summary of the invention

[0016] The purpose of the present invention is to provide an autonomous driving vehicle control method suitable for underground parking lots, so as to solve the problem that the conventional algorithms currently used in autonomous driving proposed in the above background technology cannot simultaneously meet the lane driving scenarios with small posture changes and the parking scenarios that require large posture adjustments in a small space.

[0017] To achieve the above object, the present invention provides the following technical solution: a control method for an autonomous driving vehicle suitable for an underground parking lot, comprising an obstacle avoidance control method and a vehicle entry control method, wherein both the obstacle avoidance control method and the vehicle entry control method have a vehicle speed control system, and the obstacle avoidance control method comprises the following steps:

[0018] A1: Turn on underground parking mode;

[0019] A2: The camera on the roof is raised, the radar of the whole vehicle is turned on, and the speed is reduced to 5KM / H;

[0020] A3: The camera scans the surrounding environment, and the radar scans obstacles and moving objects;

[0021] A4: Scan heat sources, scan faces, and scan obstacle shapes;

[0022] A5: Scan the moving speed of moving objects, and scan the volume of moving objects and obstacles;

[0023] A6: Avoid moving and stationary obstacles;

[0024] The vehicle entry control method comprises the following steps:

[0025] B1: Start underground parking mode;

[0026] B2: The high-position camera on the roof is raised, the radar of the whole vehicle is turned on, and the speed is reduced to 5KM / H;

[0027] B3: Scan the distance between vehicles and the distance between multiple vehicles;

[0028] B4: Determine whether there is a parking space;

[0029] B5: When there is a parking space, the vehicle slows down for the second time until it stops and the automatic parking function is switched on;

[0030] B6: No parking space, continue to scan the distance between vehicles, and scan the distance between multiple vehicles until a parking space is found;

[0031] The vehicle control algorithm comprises the following steps:

[0032] C1: The control algorithm module in the vehicle controller receives and parses the driving path data Path output by the planning module;

[0033] C2: Calculate the vehicle control preview point P based on the vehicle's real-time speed V and the average curvature radius R of the planned path Path;

[0034] C3: According to the coordinate value of the preview point P, the expected instantaneous rotation center O of the vehicle at this time is obtained, and then the expected steering angle SteerAngle of the vehicle steering wheel is calculated.

[0035] The high-position camera includes a 3D scanning camera, a thermal imaging camera and a face / eye recognition camera. The high-position camera has a night vision function. The radar includes a speed measuring radar, a distance measuring radar and a tracking radar.

[0036] In a preferred example, the present invention can be further configured as follows: in A2 and B2, a lifting mechanism is provided to adjust the height of the high-position camera.

[0037] In a preferred example, the present invention can be further configured as follows: A2 and B2 are provided with height sensors.

[0038] In a preferred example, the present invention can be further configured as follows: A1 and B1 are actively turned on by the driver.

[0039] In a preferred example, the present invention can be further configured as follows: in A2 and B2, a 360-degree rotating camera is used.

[0040] In a preferred example, the present invention can be further configured as follows: a negative feedback control system is adopted in both the obstacle avoidance control method and the vehicle positioning control method.

[0041] In a preferred example, the present invention can be further configured as follows: in B3, the existing vehicle type and vehicle size are backed up in the system.

[0042] In a preferred example, the present invention may be further configured as follows: in the A6, the vehicle avoids obstacles by pre-decelerating and turning around.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The control method of the invention does not need to consider whether the vehicle is in a lane driving scenario with a small posture change or a parking scenario that requires a large posture adjustment in a small space. It only needs to calculate based on the path data output by the planning module.

[0045] 2. When the vehicle controller of the invention adopts this control method, there is no need to switch control algorithms and parameters according to different scenarios, which can provide users with a seamless autonomous driving control experience.

[0046] 3. The invention controls the vehicle driving in the underground garage through an obstacle avoidance control method and a vehicle entry control method, adopts a three-in-one camera to monitor the surrounding environment, and uses radar to detect the surrounding environment. It can comprehensively detect heat sources, moving heat sources, obstacles and moving obstacles, and only adopts dual systems, without the need for special screening of intersections in the underground garage, reducing the complexity of information processing by the automatic driving processing system, shortening the information processing time, and doing subtraction in the processing procedure, so as to most efficiently process the information fed back by the camera and radar, thereby avoiding pedestrians who suddenly jump out to the greatest extent and improving the reliability and safety of automatic driving in the underground garage.

[0047] 4. The invention adopts a high-rise camera to patrol the surrounding environment, which can observe the parking spaces in the surrounding environment to the maximum extent, and plan the route most suitable for the vehicle size to avoid the vehicle itself from scratching other vehicles. At the same time, it uses radar to monitor the distance between multiple vehicles and quickly find a parking space of the right size, thereby avoiding trial and error, reducing the trial and error rate, and improving parking efficiency and the efficiency of finding parking spaces. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a schematic diagram of the obstacle avoidance control method of the present invention;

[0049] Figure 2 It is a principle diagram of the vehicle entry control method of the present invention;

[0050] Figure 3 A coordinate diagram of vehicle driving path data of the present invention;

[0051] Figure 4 It is a geometric relationship diagram of the vehicle coordinates and the desired steering angle of the vehicle steering wheel of the present invention. DETAILED DESCRIPTION

[0052] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0053] Example 1

[0054] See also Figure 3-4 The present invention provides an embodiment: an automatic driving vehicle control method applicable to an underground parking lot, including an obstacle avoidance control method and a vehicle entry control method, wherein the obstacle avoidance control method and the vehicle entry control method both include a vehicle control algorithm, and the vehicle control algorithm includes the following steps:

[0055] (1) The control algorithm module in the vehicle controller receives and parses the driving path data Path output by the planning module.

[0056]

[0057] Where x is the x-axis coordinate of the path point in the vehicle coordinate system, y is the y-axis coordinate of the path point in the vehicle coordinate system, and θ is the heading angle of the path point in the vehicle coordinate system;

[0058] The vehicle coordinate system takes the center of the rear axle of the vehicle as the origin and the front of the vehicle as the y-axis, as follows Figure 3 :

[0059] (2) According to the vehicle's real-time speed V and the average curvature radius R of the planned path Path, the vehicle control preview point P is calculated using the following formula:

[0060] P=(P x , P y , P θ )

[0061]

[0062] Where k, α are test parameters, and d is the distance between path points

[0063] (3) Based on the coordinate value of the preview point P, the expected instantaneous rotation center O of the vehicle is obtained, and then the expected steering angle SteerAngle of the vehicle steering wheel is calculated. The geometric relationship is as follows Figure 4 :

[0064]

[0065] In the above formula, δ is the steering gear ratio of the vehicle steering wheel, and L is the vehicle wheelbase.

[0066] Example 2

[0067] See also Figure 1-2 The present invention provides an embodiment: an automatic driving vehicle control method applicable to an underground parking lot, including an obstacle avoidance control method and a vehicle entry control method, wherein the obstacle avoidance control method and the vehicle entry control method both have a vehicle speed control system, and include the following steps:

[0068] S1: Start underground parking mode;

[0069] S2: The high-position camera on the roof is raised, the radar of the whole vehicle is turned on, and the speed is reduced to 5KM / H;

[0070] S3: The camera scans the surrounding environment, the radar scans obstacles and moving objects, the size and distance of vehicles, and the distance between multiple vehicles;

[0071] S4: The camera and radar search for parking spaces and determine whether there are any;

[0072] S5: Plan a route and drive to the parking space;

[0073] S4: The camera scans the heat source, the face, and the shape of the obstacle;

[0074] S5: The radar scans the moving speed of the moving object and the volume of the moving object and obstacles;

[0075] S6: The vehicle avoids moving and non-moving obstacles;

[0076] S7: When there is a parking space available, the vehicle decelerates a second time until it stops and switches to the automatic parking function.

[0077] The high-position cameras include 3D scanning cameras, thermal imaging cameras and face / eye recognition cameras. The high-position cameras have night vision functions. The radars include speed measurement radars, ranging radars and tracking radars.

[0078] A lifting mechanism is set under the high-position camera to adjust the height of the high-position camera. When the underground parking mode is not turned on, the high-position camera falls back to the roof and does not affect the height of the vehicle.

[0079] A height sensor is set in front of the high-position camera to avoid height restrictions. The camera can be actively lowered at the height limit to prevent damage to the camera.

[0080] A1 and B1 are activated by the driver.

[0081] The high-position camera can rotate 360 ​​degrees.

[0082] Negative feedback control systems are used in both the obstacle avoidance control method and the vehicle positioning control method.

[0083] Existing vehicle types and vehicle sizes are backed up in the system, allowing for quick screening of available parking spaces.

[0084] The vehicle avoids obstacles by pre-decelerating and turning around them.

[0085] Example 3

[0086] See also Figure 1-2 The present invention provides an embodiment: an automatic driving vehicle control method applicable to an underground parking lot, including an obstacle avoidance control method and a vehicle entry control method, wherein the obstacle avoidance control method and the vehicle entry control method both have a vehicle speed control system, and include the following steps:

[0087] S1: Start underground parking mode;

[0088] S2: The high-position camera on the roof is raised, the radar of the whole vehicle is turned on, and the speed is reduced to 5KM / H;

[0089] S3: The camera scans the surrounding environment, the radar scans obstacles and moving objects, the size and distance of vehicles, and the distance between multiple vehicles;

[0090] S4: The camera and radar search for parking spaces and determine whether there are any;

[0091] S5: Plan a route and drive to the parking space;

[0092] S4: The camera scans the heat source, the face, and the shape of the obstacle;

[0093] S5: The radar scans the moving speed of the moving object and the volume of the moving object and obstacles;

[0094] S6: The vehicle avoids moving and non-moving obstacles;

[0095] S7: No parking space available, the system returns to step S4 until a parking space is found.

[0096] The high-position cameras include 3D scanning cameras, thermal imaging cameras and face / eye recognition cameras. The high-position cameras have night vision functions. The radars include speed measurement radars, ranging radars and tracking radars.

[0097] A lifting mechanism is set under the high-position camera to adjust the height of the high-position camera. When the underground parking mode is not turned on, the high-position camera falls back to the roof and does not affect the height of the vehicle.

[0098] A height sensor is set in front of the high-position camera to avoid height restrictions. The camera can be actively lowered at the height limit to prevent damage to the camera.

[0099] A1 and B1 are activated by the driver.

[0100] The high-position camera can rotate 360 ​​degrees.

[0101] Negative feedback control systems are used in both the obstacle avoidance control method and the vehicle positioning control method.

[0102] Existing vehicle types and vehicle sizes are backed up in the system, allowing for quick screening of available parking spaces.

[0103] The vehicle avoids obstacles by pre-decelerating and turning around them.

[0104] Example 4

[0105] See also Figure 1-2 The present invention provides an embodiment: an automatic driving vehicle control method applicable to an underground parking lot, including an obstacle avoidance control method and a vehicle entry control method, wherein the obstacle avoidance control method and the vehicle entry control method both have a vehicle speed control system, and include the following steps:

[0106] S1: Start underground parking mode;

[0107] S2: The high-position camera on the roof is raised, the radar of the whole vehicle is turned on, and the speed is reduced to 5KM / H;

[0108] S3: The camera scans the surrounding environment, the radar scans obstacles and moving objects, the size and distance of vehicles, and the distance between multiple vehicles;

[0109] S4: The camera and radar search for parking spaces and determine whether there are any;

[0110] S5: Plan a route and drive to the parking space;

[0111] S4: The camera scans the heat source, the face, and the shape of the obstacle;

[0112] S5: The radar scans the moving speed of the moving object and the volume of the moving object and obstacles;

[0113] S6: The vehicle avoids moving and non-moving obstacles;

[0114] S7: There is a parking space but no parking space is available, the system returns to step S4 until a parking space is found.

[0115] The high-position cameras include 3D scanning cameras, thermal imaging cameras and face / eye recognition cameras. The high-position cameras have night vision functions. The radars include speed measurement radars, ranging radars and tracking radars.

[0116] A lifting mechanism is set under the high-position camera to adjust the height of the high-position camera. When the underground parking mode is not turned on, the high-position camera falls back to the roof and does not affect the height of the vehicle.

[0117] A height sensor is set in front of the high-position camera to avoid height restrictions. The camera can be actively lowered at the height limit to prevent damage to the camera.

[0118] A1 and B1 are activated by the driver.

[0119] The high-position camera can rotate 360 ​​degrees.

[0120] Negative feedback control systems are used in both the obstacle avoidance control method and the vehicle positioning control method.

[0121] Existing vehicle types and vehicle sizes are backed up in the system, allowing for quick screening of available parking spaces.

[0122] The vehicle avoids obstacles by pre-decelerating and turning around them.

[0123] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for controlling an autonomous driving vehicle in an underground parking lot, including an obstacle avoidance control method and a vehicle entry control method, characterized in that: The obstacle avoidance control method and the vehicle entry control method both have a vehicle speed control system, and both the obstacle avoidance control method and the vehicle entry control method include a vehicle control algorithm. The obstacle avoidance control method includes the following steps: A1: Turn on underground parking mode; A2: The camera on the roof is raised, the radar of the whole vehicle is turned on, and the speed is reduced to 5KM / H; A3: The camera scans the surrounding environment, and the radar scans obstacles and moving objects; A4: Scan heat sources, scan faces, and scan obstacle shapes; A5: Scan the moving speed of moving objects, and scan the volume of moving objects and obstacles; A6: Avoid moving and stationary obstacles; The vehicle entry control method comprises the following steps: B1: Start underground parking mode; B2: The high-position camera on the roof is raised, the radar of the whole vehicle is turned on, and the speed is reduced to 5KM / H; B3: Scan the distance between vehicles and the distance between multiple vehicles; B4: Determine whether there is a parking space; B5: When there is a parking space, the vehicle slows down for the second time until it stops and the automatic parking function is switched on; B6: No parking space, continue to scan the distance between vehicles, and scan the distance between multiple vehicles until a parking space is found; The vehicle control algorithm comprises the following steps: C1: The control algorithm module in the vehicle controller receives and parses the driving path data Path output by the planning module; C2: Calculate the vehicle control preview point P based on the vehicle's real-time speed V and the average curvature radius R of the planned path Path; C3: According to the coordinate value of the preview point P, the expected instantaneous rotation center O of the vehicle at this time is obtained, and then the expected steering angle SteerAngle of the vehicle steering wheel is calculated. The high-position camera includes a 3D scanning camera, a thermal imaging camera and a face / eye recognition camera. The high-position camera has a night vision function. The radar includes a speed measuring radar, a distance measuring radar and a tracking radar.

2. The method for controlling an automatic driving vehicle in an underground parking lot according to claim 1, characterized in that: In A2 and B2, a lifting mechanism is provided to adjust the height of the high-position camera.

3. The method for controlling an automatic driving vehicle in an underground parking lot according to claim 1, characterized in that: A2 and B2 are provided with height sensors.

4. The method for controlling an automatic driving vehicle in an underground parking lot according to claim 1, characterized in that: A1 and B1 are actively opened by the driver.

5. The automatic driving vehicle control method applicable to an underground parking lot according to claim 1, characterized in that: In A2 and B2, a 360-degree rotating camera is used.

6. The method for controlling an automatic driving vehicle in an underground parking lot according to claim 1, characterized in that: The obstacle avoidance control method and the vehicle positioning control method both adopt a negative feedback control system.

7. The method for controlling an automatic driving vehicle in an underground parking lot according to claim 6, characterized in that: In B3, the existing vehicle types and vehicle sizes are backed up in the system.

8. The automatic driving vehicle control method applicable to an underground parking lot according to claim 1, characterized in that: In the A6, the vehicle avoids obstacles by pre-decelerating and turning around.

Images