Vehicle

Abstract

The invention discloses a vehicle, and belongs to the technical field of vehicles. According to the vehicle, when the vehicle travels to a turning section, the steering angle of the vehicle is controlled between the starting point position and the end point position of the turning section to change in a mode including a plurality of peak values within a target frequency range, and the target frequency range ranges from five times per five seconds to ten times per five seconds. The vehicle is easy to control, smoother and dynamically adaptive steering control is achieved, frequent operation on a steering wheel is reduced, and the riding experience of passengers is improved.

Description

Technical Field

[0001] This application belongs to the field of vehicle technology, and in particular relates to a vehicle. Background Technology

[0002] During trajectory tracking, due to the differences between the vehicle's turning characteristics and the ideal model, as well as the response delay of the actuator, all tracking algorithms need to perform special processing on the selection of trajectory points given in the planning. When selecting tracking points, a forward look distance is usually introduced to balance tracking accuracy and tracking stability.

[0003] Currently, the forward viewpoint is usually selected on the planned trajectory. When the forward view distance is too large, the vehicle's deviation from the predetermined trajectory is large; when the forward view distance is too small, the vehicle's steering wheel operation is frequent. Summary of the Invention

[0004] This application aims to address at least one of the technical problems existing in the prior art. To this end, this application proposes a vehicle that is simple to control, achieves smoother and more dynamically adaptive steering control, reduces frequent operation of the steering wheel, and improves the riding experience for occupants.

[0005] In a first aspect, this application provides a vehicle in which, when the vehicle travels to a turning segment, the steering angle of the vehicle is controlled to change within a target frequency range in the form of multiple peaks between the starting position and the ending position of the turning segment, and the target frequency range is between 5 times per 5 seconds and 10 times per 5 seconds.

[0006] According to the vehicle provided in the embodiments of this application, by controlling the vehicle's steering angle within a moderate target frequency range during the turning segment, and varying it in the form of multiple peaks, the error between the vehicle's actual driving trajectory and the planned trajectory can be reduced. The control is simple, achieving smoother and more dynamically adaptive steering control, reducing the frequent operation of the steering wheel, and improving the passenger's riding experience.

[0007] In one embodiment of this application, the vehicle has a driving speed of any value between 3 km / h and 5 km / h, and the turning radius of the turning segment is any value between 5 m and 8 m.

[0008] In one embodiment of this application, the vehicle is a parking section. When the vehicle is parking and the predicted turning angle of the vehicle is greater than a preset angle, the steering angle of the vehicle varies in the form of multiple peaks within a target frequency range.

[0009] In one embodiment of this application, the vehicle has an error between its actual driving trajectory and its planned trajectory that is within a target error range, wherein the target error range is between -15cm and 25cm, and the actual driving trajectory is the trajectory obtained by the vehicle driving with the current driving parameters.

[0010] In one embodiment of this application, the target error range of the vehicle is between -10cm and 10cm from 5 seconds after the vehicle passes the starting position to 5 seconds before the vehicle passes the ending position.

[0011] In one embodiment of this application, the current driving parameters of a vehicle are determined based on a first tracking position. At a certain moment, the first tracking position and the center of curvature of the planned trajectory are located on opposite sides of the planned trajectory.

[0012] One embodiment of the present application includes a vehicle with a rear axle, wherein, in a top view, the straight-line distance between the center of the rear axle and the first tracking position satisfies the target spacing range.

[0013] In one embodiment of this application, the target spacing of the vehicle is between 2m and 12m.

[0014] According to one embodiment of this application, the vehicle includes a control module, which is used to determine a first tracking position. The first tracking position is determined based on a second tracking position located on the planned trajectory. The current driving parameters of the vehicle are determined based on the first tracking position. The first tracking position is located on a simulated trajectory, and the simulated trajectory is the trajectory corresponding to the simulated driving of the vehicle with the current driving parameters.

[0015] In one embodiment of this application, a vehicle aims to minimize the deviation between the simulated trajectory and the planned trajectory, and determines the first tracking position while satisfying the vehicle's driving constraints.

[0016] Secondly, this application provides a vehicle in which, when the vehicle is traveling on a turning segment, current driving parameters are determined based on a first tracking position. At a certain moment, the first tracking position and the center of curvature of the planned trajectory of the vehicle are located on opposite sides of the planned trajectory.

[0017] One embodiment of the present application includes a vehicle with a rear axle, wherein, in a top view, the straight-line distance between the center of the rear axle and the first tracking position satisfies the target spacing range.

[0018] According to one embodiment of this application, the vehicle includes a control module, which is used to determine a first tracking position. The first tracking position is determined based on a second tracking position located on the planned trajectory. The current driving parameters of the vehicle are determined based on the first tracking position. The first tracking position is located on a simulated trajectory, and the simulated trajectory is the trajectory corresponding to the simulated driving of the vehicle with the current driving parameters.

[0019] In one embodiment of this application, a vehicle aims to minimize the deviation between the simulated trajectory and the planned trajectory, and determines the first tracking position while satisfying the vehicle's driving constraints.

[0020] In one embodiment of this application, the vehicle's driving constraints include at least one of the following: the simulated trajectory is within a target driving range, the vehicle's speed is within a target speed range, the vehicle's acceleration is within a target acceleration range, and the curvature of the vehicle's simulated trajectory is within a target curvature range.

[0021] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0022] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a schematic diagram of the vehicle structure provided in the embodiments of this application; Figure 2 This is a schematic diagram of the test scenario provided in the embodiments of this application; Figure 3 This is one of the schematic diagrams of the test results provided in the embodiments of this application; Figure 4 This is a second schematic diagram of the test results provided in the embodiments of this application; Figure 5 This is the third schematic diagram of the test results provided in the embodiments of this application; Figure 6 This is the fourth schematic diagram of the test results provided in the embodiments of this application; Figure 7 This is the fifth schematic diagram of the test results provided in the embodiments of this application; Figure 8 This is the sixth schematic diagram of the test results provided in the embodiments of this application; Figure 9 This is the seventh schematic diagram of the test results provided in the embodiments of this application; Figure 10 This is the eighth schematic diagram of the test results provided in the embodiments of this application; Figure 11 This is the ninth schematic diagram of the test results provided in the embodiments of this application; Figure 12 This is a schematic diagram illustrating the tracking effect of the vehicle's driving trajectory versus the planned trajectory under ideal conditions in related technologies; Figure 13 This is a schematic diagram illustrating the effect of tracking vehicle trajectory versus planned trajectory in actual situations using related technologies. Figure 14 This is one of the schematic diagrams illustrating the effective range of trigonometric functions and the trend of error changes after failure in related technologies; Figure 15 This is the second illustration of the effective range of trigonometric functions in related technologies and the trend of error changes after failure; Figure 16 This is the third illustration of the effective range of trigonometric functions in related technologies and the trend of error changes after failure; Figure 17 This is the fourth illustration of the effective range of trigonometric functions in related technologies and the trend of error changes after failure; Figure 18 This is the fifth illustration of the effective range of trigonometric functions in related technologies and the trend of error changes after failure; Figure 19 This is the sixth illustration of the effective range of trigonometric functions in related technologies and the trend of error changes after failure; Figure 20 This is a schematic diagram illustrating the trajectory difference between a vehicle understeering and its planned trajectory in related technologies; Figure 21 This is a schematic diagram illustrating the effect of correcting trajectory differences in related technologies; Figure 22 This is a schematic diagram illustrating the effect of correcting trajectory differences provided in the embodiments of this application; Figure 23 This is one of the flowcharts illustrating the vehicle control method provided in the embodiments of this application; Figure 24 This is a second schematic flowchart of the vehicle control method provided in the embodiments of this application; Figure 25 This is a schematic diagram of the vehicle turning radius fitting effect provided in the embodiments of this application.

[0023] Figure label: Vehicle 110, control module 120. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0025] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0026] The vehicle provided in the embodiments of this application will be described in detail below with reference to the accompanying drawings and through specific implementation methods and application scenarios.

[0027] The vehicle is used to pass through a U-shaped road test, which includes controlling the vehicle to drive in a U-shaped road condition. The U-shaped road condition includes a first straight driving section, a turning section, and a second straight driving section set in sequence. When controlling the vehicle to drive in the turning section, the vehicle reaches the starting position of the turning section, which corresponds to the first moment, and the vehicle reaches the ending position of the turning section, which corresponds to the second moment.

[0028] It is understandable that the first straight section connects to one end of the turning section, and the other end of the turning section connects to the second straight section.

[0029] The vehicle is connected to a control module, which controls the vehicle's steering angle to vary within a target frequency range, including multiple peaks, between the start and end positions of the turning segment when the vehicle is traveling to a turning segment.

[0030] When the vehicle is traveling on the first straight section, the vehicle's steering angle is basically 0 (i.e., the steering wheel remains in a straight position with no obvious steering action; for example, if the difference between the vehicle's steering angle and 0 is less than 2 degrees or other values, the vehicle's steering angle can be considered to be basically 0). As the vehicle is about to leave the first straight section, the vehicle's steering angle begins to gradually increase (i.e., the steering wheel begins to turn in the direction of the turn, officially entering the turning posture). That is, the position corresponding to the moment when the steering angle is increased from basically 0 to a first preset steering angle (such as 15 degrees or 20 degrees, or other values) and maintained for a first duration (such as 1 second or 2 seconds, or other values) is determined as the starting position of the turning segment.

[0031] When a vehicle is driving on a curve, its steering angle is constantly changing. As the vehicle is about to leave the curve, its steering angle gradually decreases and returns to a near-zero state (steering wheel straightens, preparing to enter the second straight driving segment). That is, the position corresponding to the moment when the steering angle is reduced from a non-zero value (such as 20 degrees, or other values) to near-zero and maintained for a second duration (such as 0.5s or 1s, or other values) is determined as the end point of the curve.

[0032] The change in vehicle steering angle can be characterized by the number of times the steering wheel turns in different directions. The number of times the steering wheel turns in different directions is the number of times the steering wheel turns from clockwise to counterclockwise, or from counterclockwise to clockwise. Each complete "clockwise → counterclockwise" or "counterclockwise → clockwise" transition is counted as one transition (i.e., one change in steering angle).

[0033] The target frequency range is a pre-set range. When the frequency of steering wheel rotation direction switching is within the target frequency range, it indicates that the frequency of steering wheel rotation direction switching is low. The steering wheel rotation direction switching frequency is the number of times the steering wheel rotation direction is switched repeatedly within a unit of time (e.g., 5 seconds).

[0034] It can control the vehicle to switch the steering wheel direction multiple times between the starting and ending points of a turning segment, according to the target frequency range, to form multiple peaks. The number of switching can reflect the change pattern of the steering angle.

[0035] In this embodiment, the vehicle travels sequentially through a first straight-line segment, a turning segment, and a second straight-line segment. On the first straight-line segment before entering the turn, the vehicle's steering angle remains fixed (a first steering angle, which can be approximately 0) until the start of the turn. From the start of the turn (first moment) to the completion of the turn (second moment), the control module dynamically adjusts the steering wheel angle, ensuring that its changes include multiple peaks. This maintains the steering wheel rotation frequency within a target range, thereby reducing the error between the vehicle's actual trajectory and the planned trajectory, allowing the vehicle to travel along the preset planned trajectory as closely as possible.

[0036] The actual driving trajectory is the trajectory obtained by the vehicle in actual driving with the current driving parameters. In actual execution, there will be errors due to interference from factors such as the actual site. The planned trajectory is the driving trajectory expected by the vehicle.

[0037] The planned trajectory can be generated using a path planning algorithm, taking into account the vehicle's size, the geometry of the parking space, and surrounding obstacles.

[0038] The frequency range of the vehicle's steering angle change between the first moment and the second moment (i.e., the target frequency range) is 5 times to 10 times every 5 seconds. For example, the frequency of the vehicle's steering angle change between the first moment and the second moment can be 5 times every 5 seconds, 8 times every 5 seconds, or 10 times every 5 seconds, or it can be other values ​​within the target frequency range. This application does not limit this.

[0039] During the research and development process, the inventors discovered that the tracking algorithms for achieving automatic parking functions in related technologies mainly include geometry-based tracking algorithms and control law-based tracking algorithms. However, during trajectory tracking, due to differences between the vehicle's turning characteristics and the ideal model, as well as actuator response delays, tracking algorithms typically require special processing of the selected trajectory points. Otherwise, the tracking error may increase during large-angle turns or the tracking system may experience singular perturbations, i.e., the vehicle may follow an S-shaped trajectory during tracking. Both increased tracking error and system singular perturbations have a significant impact on parking safety and effectiveness. To avoid system singular perturbations and reduce tracking error simultaneously, the controller typically introduces a lookforward distance (LD) when selecting tracking points. The selection of the lookforward distance balances tracking accuracy and tracking stability.

[0040] Since the curvature of a straight path changes little, the control system can approximate the straight-line trajectory tracking problem as a constant tracking problem. However, during the entry, exit, and turning of a vehicle at a large angle, due to the sudden change in road curvature and errors in vehicle model assumptions, the tracking algorithm in related technologies will lead to a worse tracking effect, a larger lateral error in trajectory tracking, and lower stability due to the back-and-forth shaking of the steering wheel.

[0041] During the research and development process, the inventors discovered that in related technologies, in order to reduce the error between the actual driving trajectory and the planned trajectory, high-frequency and high-gain closed-loop feedback control is usually used. For example, driving parameters (including steering angle) are adjusted at high frequency to reduce the error between the actual driving trajectory and the planned trajectory. However, the control schemes in related technologies are complex and have poor driving stability. In scenarios with large-angle turns, such as automatic parking, the curvature of the planned trajectory changes drastically. The high-frequency and high-gain controller will react violently to this, causing frequent and large-amplitude oscillations in the steering angle command. This will cause the vehicle steering wheel to vibrate violently and the vehicle body to sway from side to side, reducing the passenger riding experience and even causing vehicle instability.

[0042] In this application, the steering angle is controlled to vary in the form of multiple peaks within a target frequency range (i.e., 5 times per second to 10 times per 5 seconds). The target frequency range is a relatively low-frequency adjustment range, which is in a stable and effective range in driving feel. The control is simple, the error between the actual driving trajectory and the planned trajectory is reduced, and the riding experience of the occupants is improved.

[0043] According to the vehicle provided in the embodiments of this application, by controlling the vehicle's steering angle within a moderate target frequency range during the turning segment, and varying it in the form of multiple peaks, the error between the vehicle's actual driving trajectory and the planned trajectory can be reduced. The control is simple, achieving smoother and more dynamically adaptive steering control, reducing the frequent operation of the steering wheel, and improving the passenger's riding experience.

[0044] In some embodiments, the vehicle speed is any value between 3 km / h and 5 km / h, and the turning radius of the turning section is any value between 5 m and 8 m.

[0045] In this embodiment, the vehicle's speed in the U-shaped road condition can be any value between 3 km / h and 5 km / h (e.g., the vehicle's speed can be 3 km / h, 4 km / h, or 5 km / h, or other values, which are not limited in this application). The length of the first straight driving segment can be between 5 m and 15 m (e.g., the length of the first straight driving segment can be 5 m, 10 m, or 15 m, or other values, which are not limited in this application). The turning radius of the turning segment that satisfies the maximum turning angle of the vehicle can be any value between 5 m and 8 m (e.g., 5 m, 6 m, 8 ... The length of the first straight driving segment can be 5m to 15m (e.g., the length of the second straight driving segment can be 5m, 10m or 15m, or other values, which are not limited in this application). The first straight driving segment is parallel to the second straight driving segment, and the distance between the first straight driving segment and the second straight driving segment can be 10m to 16m (e.g., the distance between the first straight driving segment and the second straight driving segment can be 10m, 13m or 16m, or other values, which are not limited in this application).

[0046] In some embodiments, the turning segment is a parking segment, and when the vehicle is parking and the predicted turning angle of the vehicle is greater than a preset angle, the steering angle of the vehicle varies in the form of multiple peaks within the target frequency range.

[0047] In this embodiment, the turning segment is the trajectory corresponding to the vehicle parking, that is, the predetermined driving path of the vehicle from the current position to the final parking space. The planned trajectory can ensure that the vehicle can smoothly and safely enter the parking space.

[0048] The predicted turning angle is the angle that the vehicle is likely to turn.

[0049] The predicted turning angle is the maximum steering angle (or the steering angle at key nodes) that the steering system may need to reach when the vehicle is driving along the planned trajectory during parking.

[0050] The predicted turning angle can be derived from the radius of curvature of the planned trajectory, or it can be obtained by fitting historical parking data, or it can be obtained by other means, which are not limited in this application.

[0051] During the parking process, the steering angle of the vehicle can be adjusted based on the predicted turning angle or changes in the vehicle's turning angle, depending on the first tracking position.

[0052] In some embodiments, the control module is configured to adjust the vehicle's steering angle based on a first tracking position when the vehicle is parking and the predicted turning angle of the vehicle is greater than a preset angle.

[0053] The preset angle is a pre-set angle value that can be determined based on factors such as the vehicle model.

[0054] In this embodiment, if the predicted turning angle of the vehicle is greater than the preset angle, it indicates that the vehicle may make a large-angle turn. In the case that the vehicle may make a large-angle turn, the steering angle of the vehicle is adjusted based on the first tracking position.

[0055] In some embodiments, the control module is used to adjust the vehicle's steering angle based on the vehicle's second tracking position when the vehicle is parking and the vehicle's predicted turning angle is less than or equal to a preset angle, or when the vehicle is in motion. The second tracking position is on the planned trajectory.

[0056] Among them, the driving state refers to the state in which the vehicle is not parked.

[0057] In this embodiment, if the predicted turning angle of the vehicle is less than or equal to a preset angle, it indicates that the vehicle may make a small-angle turn. When the vehicle is parked and making a small-angle turn or when the vehicle is not parked, the steering angle of the vehicle is adjusted based on the second tracking position of the vehicle, and the vehicle is controlled to move towards the second tracking position. By tracking the second tracking position, the vehicle can move along the planned trajectory.

[0058] In some embodiments, the error between the vehicle's actual driving trajectory and the planned trajectory is within the target error range.

[0059] In this embodiment, the error between the vehicle's actual driving trajectory and the planned trajectory can be understood as the lateral error, which is the distance from the vehicle's current position to the nearest point on the planned trajectory.

[0060] The target error range is between -15cm and 25cm (e.g., it can be -15cm, -5cm, 0cm, 15cm or 25cm, etc.). By controlling the vehicle's steering angle, the tracking error during the entire turning process can be stably controlled within the target error range.

[0061] If the error between the vehicle's actual trajectory and the planned trajectory is -15cm, it indicates that the vehicle is 15cm inside the planned trajectory. If the error between the vehicle's actual trajectory and the planned trajectory is 25cm, it indicates that the vehicle is 25cm outside the planned trajectory.

[0062] The actual driving trajectory is the trajectory obtained by the vehicle traveling with the current driving parameters, which may include the vehicle's current speed and acceleration.

[0063] In some embodiments, the target error range is between -10cm and 10cm between 5 seconds after the vehicle passes the starting position and 5 seconds before it passes the ending position.

[0064] In this embodiment, the period from when the vehicle passes the starting point to when it enters the turning segment (5 seconds later) may be the adjustment phase when the vehicle is just entering the turn; the period from when the vehicle passes the end point (5 seconds before) to when it reaches the end point may be the adjustment phase when the vehicle is about to exit the turn.

[0065] The section of road between 5 seconds after the vehicle passes the starting point and 5 seconds before the ending point is the core and stable turning section. Within this turning section, the error between the vehicle's actual trajectory and the planned trajectory can be minimized. That is, in the main part of the curve center, this control method can make the error between the vehicle's actual trajectory and the planned trajectory between -10cm and 10cm (e.g., -10cm, -5cm, 0cm, 5cm, or 10cm, etc.).

[0066] In some embodiments, at a given moment, the first tracking position and the center of curvature of the planned trajectory are located on opposite sides of the planned trajectory.

[0067] In some embodiments, the vehicle includes a rear axle, and in a top-view direction, the straight-line distance from the center of the rear axle to the first tracking position satisfies the target spacing range.

[0068] In this embodiment, when the vehicle is traveling on a turning section, at a certain moment, the first tracking position is located on the outward convex side of the planned trajectory, and the distance between the first tracking position and the center of the vehicle's rear wheel axle is within the target distance range.

[0069] The control module is used to adjust the vehicle's steering angle based on the first tracking position.

[0070] The first tracking position is the position tracked during vehicle movement. The first tracking position is located on the outward protruding side of the target driving trajectory, i.e., the outer circle of the turning direction. The distance between the first tracking position and the center of the rear wheel axle of the vehicle is within the target distance range. The target distance range can be set based on the vehicle's driving speed. For example, during parking, if the vehicle's driving speed is between 3 km / h and 20 km / h (e.g., it can be 3 km / h, 10 km / h, or 20 km / h, etc., or any value between 3 km / h and 20 km / h, this application does not limit it), the corresponding target distance range can be 2m-12m (e.g., it can be 2m, 6m, 10m, or 12m, etc.).

[0071] For example, in scenarios where the vehicle speed is 3km / h-8km / h (such as low-speed parking scenarios where the vehicle moves slowly in a narrow space), the corresponding target distance range can be 2m-5m. The distance is relatively short, which can be adapted to precise steering tracking at low speeds. For example, when the vehicle speed is 5km / h, the corresponding target distance range can be 3.5m. In scenarios where the vehicle speed is 8km / h-15km / h (medium-speed parking scenario, such as moving a car in an open parking lot), the corresponding target distance range can be 5m-9m. The distance is moderate and can balance steering flexibility and tracking stability. For example, when the vehicle speed is 11km / h, the corresponding target distance range can be 7m. In scenarios where the vehicle speed is 15km / h-20km / h (high-speed parking scenarios, such as long-distance parking space connections), the corresponding target spacing range can be 9m-12m. The spacing is relatively long, which can adapt to steering prediction and trajectory matching at high speeds. For example, when the vehicle speed is 18km / h, the corresponding target spacing range can be 11m.

[0072] The control module directs the vehicle toward the first tracking position, but does not actually reach the first tracking position. The process of directing the vehicle toward the first tracking position corresponds to the process of adjusting the vehicle's steering angle, so that the vehicle will not cut too inward, resulting in an excessively small trajectory, nor will it deviate from the outside due to understeer. This ensures that the vehicle travels approximately along the target trajectory, achieving a smooth, safe, and controllable turning process.

[0073] In some embodiments, the vehicle includes a control module for determining a first tracking position.

[0074] In this embodiment, the first tracking position can be determined based on the second tracking position, wherein the second tracking position is located on the planned trajectory, such as... Figure 22 As shown, D' is the first tracking position, D is the second tracking position, and OD is the planned trajectory.

[0075] The vehicle's current driving parameters can be determined based on the first tracking position.

[0076] Without considering interference from factors such as the actual terrain, the vehicle simulates driving using the current driving parameters, resulting in a simulated trajectory. The first tracking position is located on this simulated trajectory, such as... Figure 22 As mentioned above, OD' is the simulated trajectory.

[0077] In some embodiments, the first tracking position is determined with the goal of minimizing the deviation between the simulated trajectory and the planned trajectory, while satisfying the vehicle's driving constraints.

[0078] In this embodiment, the driving constraints are conditions that constrain the vehicle's physical performance, safe distance, etc.

[0079] An objective function can be defined to quantify the deviation between the simulated trajectory and the planned trajectory.

[0080] Optimization algorithms such as gradient descent can be selected to find the minimum value of the objective function while satisfying driving constraints, in order to obtain the optimal tracking position.

[0081] In some embodiments, the driving constraints include at least one of the following: the simulated trajectory is within a target driving range, the vehicle speed is within a target speed range, the vehicle acceleration is within a target acceleration range, and the curvature of the simulated trajectory is within a target curvature range.

[0082] In this embodiment, the target driving range is the area that the vehicle needs to stay in during actual driving, which can be determined according to the specific application scenario and safety requirements.

[0083] The target speed range is the range between the maximum speed limit that a vehicle is not allowed to exceed while driving and the minimum speed at which a vehicle can normally make a turn. It can be set based on road type and traffic regulations.

[0084] The target acceleration range is the range between the maximum acceleration limit that the vehicle is not allowed to exceed during driving and the minimum acceleration that the vehicle can achieve normal turning. It can be set based on the vehicle's physical characteristics and passenger comfort requirements.

[0085] The target curvature range is the curvature corresponding to the minimum turning radius that the vehicle is not allowed to fall below when turning. The larger the curvature, the smaller the turning radius. The preset curvature threshold can be set based on the vehicle's physical characteristics and road conditions.

[0086] The minimum turning radius of a vehicle can be calculated using the following formula.

[0087]

[0088]

[0089] in, The wheelbase between the front and rear wheels of the vehicle. The angle between the midpoint of the rear axle and the front wheel steering angle. This is the line connecting the start and end points of the vehicle's turning arc. For the turning angle of the wheel, This represents the vehicle's current turning radius.

[0090] The specific values ​​of the target speed range, target acceleration range, and target curvature range can be determined by combining the vehicle model parameters, engine parameters, vehicle steering mechanism configuration, and actual global planning trajectory.

[0091] In this embodiment, driving within the target driving range ensures that the vehicle does not enter dangerous areas or violate traffic rules.

[0092] The vehicle's speed is within the target speed range to ensure the safety of passengers and pedestrians near the vehicle.

[0093] When a vehicle's acceleration is within the target speed range, passenger comfort can be guaranteed, avoiding discomfort caused by excessive acceleration or deceleration, while also preventing the vehicle from losing control due to excessive inertial forces.

[0094] The curvature of the vehicle's simulated trajectory, within the target curvature range, ensures the vehicle's stability and safety when turning.

[0095] In actual execution, an ideal planned trajectory can be obtained, and the part with greater curvature in the planned trajectory is identified as the turning segment. Under certain conditions (such as the predicted turning angle of the vehicle being greater than the preset angle and the driving speed being between 3km / h and 5km / h), the system determines that the core control logic needs to be activated.

[0096] A second tracking position can be selected on the planned trajectory. The control module can simulate the trajectory that the vehicle will take if it directly tracks the second tracking position, based on the vehicle's current driving parameters.

[0097] The control module can minimize the deviation between the simulated trajectory and the planned trajectory (e.g., it can obtain the area of ​​the closed shape formed by OD, OD', and DD' and determine it as the deviation between the simulated trajectory and the planned trajectory) and find a better tracking point.

[0098] In the process of finding a better tracking point, it is necessary to constrain the vehicle to meet driving constraints, such as limiting the vehicle's simulated trajectory to within the target driving range, limiting the vehicle's speed and acceleration to within the target speed and acceleration ranges, and limiting the curvature of the simulated trajectory to within the target curvature range (i.e., not exceeding the vehicle's minimum turning capability).

[0099] Under the premise of satisfying all the above driving constraints, the control module can calculate and determine an optimal point on the simulated trajectory, namely the first tracking position. The first tracking position and the center of curvature of the planned trajectory are located on opposite sides of the planned trajectory, that is, the first tracking position is located outside the planned trajectory.

[0100] The control module can switch the vehicle's tracking target from the second tracking position to the first tracking position. Based on the first tracking position, the control module can recalculate a new set of current driving parameters.

[0101] The controller can repeat the above process throughout the entire cornering process, from the first moment to the second moment.

[0102] By continuously tracking the dynamically changing first tracking position, the vehicle's actual driving trajectory can gradually approach the planned trajectory. The vehicle's steering angle will change in the form of multiple peaks within the target frequency range, and the error between the vehicle's actual driving trajectory and the planned trajectory can be controlled within the target error range.

[0103] In this application, a constrained optimization process is used to dynamically correct the tracking point (point D) from the planned trajectory to a new position (point D') outside the planned trajectory, thereby enabling guidance and control of the vehicle's driving direction to achieve smooth automatic parking.

[0104] In this embodiment, based on the current position, the second tracking position, and the current driving parameters, the vehicle's driving path over a future period is predicted using a vehicle dynamics model, thereby determining the simulated trajectory. The second tracking position is a location on the planned trajectory that is relatively close to the vehicle.

[0105] In this embodiment, the planned trajectory is the desired path trajectory of the vehicle, and the simulated trajectory is the path trajectory that the vehicle is expected to travel under the current driving parameters.

[0106] It should be noted that the planned trajectory can be changed during vehicle operation. For example, if a pedestrian is detected near the vehicle while it is in motion, the system can automatically adjust the planned trajectory to avoid the pedestrian. The driver can also manually trigger the adjustment of the planned trajectory based on the actual road conditions seen by the driver. The planned trajectory is updated after manual triggering.

[0107] In this step, the vehicle's control system can automatically generate a planned trajectory based on the vehicle's destination and the surrounding road conditions.

[0108] In this embodiment, the simulated trajectory is compared with the planned trajectory, the positional deviation between the two is analyzed, and based on the analysis results, the vehicle is guided to approach the planned trajectory. The vehicle's driving situation is simulated through a dynamic vehicle model to generate the first tracking position.

[0109] It should be noted that the first tracking location is updated in real time while the vehicle is in motion.

[0110] It should be noted that the first tracking position is not the position in the planned trajectory. Through continuous correction, the initial segment of the vehicle's actual driving trajectory approximately coincides with the planned trajectory. The vehicle can drive towards the first tracking position so that it passes through a segment of the trajectory that approximately coincides with the planned trajectory.

[0111] The vehicle moves toward the first tracking position, but does not actually reach the first tracking position. The first tracking position is updated in real time, and the vehicle is always in the state of tracking the current first tracking position, so that the vehicle travels approximately along the planned trajectory throughout the entire journey.

[0112] In this embodiment, the first tracking position of the vehicle is obtained based on the planned trajectory and the simulated trajectory, and the vehicle is controlled to drive toward the first tracking position, but does not actually reach the first tracking position. The first tracking position is updated in real time, and the vehicle is always in the state of tracking the current first tracking position, so that the vehicle can drive approximately along the planned trajectory throughout the entire process. Based on the real-time updated first tracking position, the vehicle can achieve smoother and more dynamically adaptive steering control, thereby reducing the frequency of steering wheel operation.

[0113] In some embodiments, the control module is used to determine a first tracking position with the objective of minimizing the deviation between the simulated trajectory and the planned trajectory, while satisfying the vehicle's driving constraints.

[0114] In some embodiments, the control module is further configured to obtain the current geometric parameters corresponding to the vehicle turning; And control the vehicle's movement using the current geometric parameters.

[0115] Among them, the geometric parameters corresponding to vehicle turning include wheelbase, track width and steering angle. The geometric parameters corresponding to vehicle turning affect the turning performance and stability of the vehicle. The current geometric parameters are the geometric parameters corresponding to the vehicle at the current moment.

[0116] In this embodiment, before controlling the vehicle to drive, the geometric parameters corresponding to the vehicle turning are calibrated to obtain the current geometric parameters. By controlling the vehicle to drive using the current geometric parameters, the behavior of the vehicle during the turning process can be made to meet expectations, thereby improving the accuracy and safety of driving.

[0117] In some embodiments, the current geometry parameters include at least the vehicle’s current turning radius.

[0118] The turning radius of a vehicle is the distance from the center of the turn to the point where the outer tire of the vehicle contacts the ground when the vehicle is turning; that is, the radius of the smallest circle drawn by the vehicle when turning.

[0119] In this embodiment, the vehicle can be controlled to turn in a fixed direction during the experiment, the actual trajectory radius of the vehicle can be measured, and then the accurate value of the turning radius can be calculated by combining mathematical methods such as the least squares method, thereby calibrating the parameters in the system.

[0120] The steering radius can also be dynamically adjusted and calibrated by collecting steering data in real time during vehicle operation and applying parameter identification algorithms.

[0121] The following uses automatic parking as an example to introduce the vehicle provided in the embodiments of this application.

[0122] It should be noted that the following methods and steps can all be executed through the control module.

[0123] The main parking scenarios include parallel parking, perpendicular parking, dead-end parking, and angled parking. Among them, parallel parking has a smaller angle and is not recommended as a test scenario. Angled parking requires a large turning angle (e.g., if the angled parking space needs to be entered from the left, the angle of entering from the right is smaller). It is also necessary to choose whether to park with the front or rear of the vehicle depending on the actual scenario. Various parking parameters need to be calculated based on the target vehicle model, which is usually 1.2-1.5 times the actual length of the vehicle.

[0124] In some embodiments, large-angle U-bends and continuous S-bends can also be used as test scenarios, and the test road conditions can be selected as standard road conditions (such as cement roads and asphalt roads), which can avoid the measurement impact caused by wheel slippage during vehicle driving due to special road conditions and extreme weather (such as rain and snow). In addition, the curvature of the large-angle turning section of the test scenario must be greater than the curvature corresponding to the minimum turning radius of the vehicle.

[0125] In this embodiment, a large-angle U-turn can be visualized. The U-turn can be in any direction. The test section can include a first straight section of 10m before entering the turn, a second large-angle turn with a turning radius of 5m that meets the maximum turning angle of the vehicle, and a third section of 10m after exiting the turn. The first and third sections are parallel, with a parallel distance of 10m, and both are tangent to the second section.

[0126] In this embodiment, all different vehicle test processes are conducted under U-shaped curve trajectory following conditions.

[0127] like Figure 2 As shown, a visual explanation is given using the example of a vehicle tracking a large-angle U-shaped bend.

[0128] The test results for right-turn U-turns are symmetrical to those for left-turn U-turns along the x-axis. This example can be illustrated using a left-turn U-turn as an example. The neutral position when the wheel is moving straight forward without deflection is assumed to be 0°. The wheel angle when the wheel rotates counterclockwise is positive (left turn), and the vehicle angle when the vehicle rotates clockwise is negative (right turn).

[0129] To balance tracking accuracy and tracking stability, such as Figures 3-11 The example illustrates the impact of different forward-looking distances (the distance between the vehicle's current position and the forward-looking point) on the tracking performance of the planned trajectory in practical applications.

[0130] like Figures 3-5 As shown, this illustrates the vehicle tracking trajectory effects under different forward-looking distances. Dark blue represents the globally planned trajectory, and red represents the actual vehicle following effect. Figures 6-8 As shown, this represents the vehicle's trajectory tracking error during driving. Figures 9-11 The image shows the change in wheel angle during driving.

[0131] The selection of forward sight distance is similar to the driver's focus when driving a vehicle. When the vehicle speed is slow and the error between the vehicle and the planned trajectory is small, the focus is usually on precise operations close to the vehicle, resulting in small but frequent steering wheel movements. When the vehicle speed is slow but the error between the vehicle and the planned trajectory is large, the priority is to get the vehicle back onto the planned trajectory, which may result in larger steering wheel movements but lower frequency. When the vehicle speed is fast and the error between the vehicle and the planned trajectory is small, the error is generally ignored and the focus is on distant objects, maintaining small steering wheel movements and fewer operations. When the vehicle speed is fast and the error is large, some error is ignored, and continuous small but frequent operations are maintained to quickly and smoothly reduce the error between the vehicle and the planned trajectory.

[0132] like Figure 3 , Figure 6 and Figure 9 As shown, when the selected forward viewing distance is too large, the controller prioritizes stability, and the response speed of the vehicle's wheel angle changes is slow. Furthermore, after reaching the threshold, in order to ensure stability, there is a steady-state lateral error of about 0.25m to avoid adjusting the steering wheel.

[0133] like Figure 4 , Figure 7 and Figure 10The diagram shows a smaller forward-looking distance. In this case, the controller focuses more on the tracking accuracy, and the accuracy fluctuation can be as low as ±0.05m. However, due to the sensitivity to error, even during large-angle turns, the steering wheel will be adjusted more frequently. When making large-angle turns, the vehicle's steering wheel will vibrate at a high frequency, and the vibration amplitude will reach about 15° (the single-side steering threshold of the wheel is generally around 35°).

[0134] The test section includes a first straight driving section of 10m before entering the curve, a large-angle curve with a turning radius of 5m, and a second straight driving section of 10m after exiting the curve. The first and second straight driving sections are parallel to each other, with a parallel distance of 10m, and both are tangent to the curve section.

[0135] like Figure 5 , Figure 8 and Figure 11 The figure shows the tracking effect of the vehicle control method provided in this application embodiment. It mainly optimizes the tracking accuracy and stability during the large-angle turning of the vehicle, so that the controller can achieve a fast response of the steering wheel but the accuracy sensitivity is reduced during large-angle turning. Even if the requirement for tracking accuracy is reduced, it can ensure an error of ±0.05m in the planned trajectory during the turning process (about 10-40s), and the frequency of steering wheel control is greatly reduced.

[0136] The vehicle travels at a constant speed of 3.6 km / h. The point of tangency between the first and second line segments is t0 (absolute time t=10s). The point where the wheel angle stabilizes after entering the curve is t1 (depending on vehicle tracking). The points equidistant from the midpoint of the second line segment are t2 (absolute time t=25s) and t3 (absolute time t=30s) on the trajectory lines before and after the midpoint of the second line segment, respectively. The point of tangency between the second and third line segments is t4 (depending on vehicle tracking). The point where the wheel angle stabilizes after exiting the curve from the second line segment to the third line segment is t5 (depending on vehicle tracking).

[0137] from Figure 3 , Figure 6 and Figure 9It can be seen that the time from entering the curve to stabilizing is t1-t0=8s, during which the wheel angle changes by 30°. The vehicle controller enters the curve relatively early, but the period from the start of the curve to stabilizing is relatively long, and the lateral error increases rapidly to 0.28m. At this time, the selected forward sight distance is too large, and the controller prioritizes stability. The response priority of the wheel angle change is increased, and the response speed is reduced. During the stabilizing curve period t4-t1=20s, in order to ensure stability, there is a steady-state lateral error of about 0.25m during the large-angle stable turn. At the midpoint of the large-angle turn t3-t2=5s, the wheel angle includes 8 reverse sinusoidal waveform changes (one reverse change of the wheel angle is counted as half a sinusoidal waveform period), with a maximum peak fluctuation of 3°, which is a stable underdamped fluctuation state. Due to the reduced response speed, it takes t5-t4=9s from exiting the curve to stabilizing.

[0138] and Figure 4 , Figure 7 and Figure 10 This approach uses a smaller forward-looking distance, with a turn-in to turn-out stabilization time of t1-t0=1.5s. During this time, the wheel angle changes by 33°. At this point, the controller focuses more on tracking accuracy, prioritizing and increasing the response speed to changes in the vehicle's wheel angle. As a result, the accuracy fluctuation is kept as low as ±0.05m, and the stabilization turn-out time of t4-t1=28s is significantly extended. However, due to its sensitivity to errors, even during the midpoint of a large-angle turn (t3-t2=5s), the steering wheel is adjusted at a high frequency. The wheel angle includes 23 sharp reverse waveform changes, and the peak waveform variation can reach 15°. In real-world driving, this manifests as high-end steering wheel vibration during large-angle turns (the single-sided wheel steering threshold is generally around 35°), which greatly affects driving safety. However, thanks to its high response, the turn-out to stabilization time is only t5-t4=2.1s.

[0139] and Figure 5 , Figure 8 and Figure 11The tracking effect of the vehicle control method provided in this application embodiment is mainly optimized for tracking accuracy and stability during large-angle steering. The controller achieves rapid steering wheel response but reduces accuracy sensitivity during large-angle steering. As can be seen from the figure, the time from entering the curve to stabilizing is t1-t0=1.8s. Even with the reduced requirement for tracking accuracy, the stable cornering period time reaches t4-t1=26s. Furthermore, during the stable cornering period, it can ensure an error of ±0.05m on the planned trajectory, which is close to the error fluctuation range of a smaller forward sight distance, but much better than the ±0.25m error of a larger forward sight distance. During the 5s period from t3 to t2 at the midpoint of a sharp turn, although the maximum peak fluctuation of the wheel angle is 6°, which is greater than the 3° fluctuation at a larger forward sight distance, it is only half of the 15° fluctuation at a smaller forward sight distance. Furthermore, it only exhibits 7 reverse sinusoidal waveform changes, close to the 8 changes at a larger forward sight distance, but far lower than the 23 changes at a smaller forward sight distance. Since the response speed is not suppressed, it still takes t5 to t4 = 2.7s to reach stability after exiting the turn. Under these conditions, the vehicle's tracking accuracy during sharp turns is guaranteed, with minimal interference to the steering wheel, thus improving safety and passenger experience.

[0140] The inventors' tests revealed that by reasonably controlling the frequency and amplitude of wheel turning angles through the vehicle control method provided in this application, the trajectory tracking accuracy, driving safety, and driving comfort of the vehicle during large-angle turning can be improved.

[0141] Next, we will analyze the sources of error introduction step by step. The main sources of steering wheel vibration during large-angle steering are as follows:

[0142] in, This is the vehicle's heading angle.

[0143] During the research and development process, the inventors discovered that low-speed parking scenarios are generally based on kinematic models. When modeling kinematic models, some simplifications can be made. For example, during planning, the vehicle can be simplified to a point with volume, or the vehicle model can be simplified to a bicycle model with only two wheels to reduce the computational load at the algorithm level. However, this will cause a difference between the model and the actual vehicle. A point can move directly along an arc, but the vehicle's steering is understeering. Therefore, the actual trajectory of the vehicle will not be an ideal arc trajectory.

[0144] like Figure 12 As shown, this illustrates the tracking effect between the ideal driving trajectory and the planned trajectory of a vehicle under ideal conditions (the vehicle is simplified to a point). Figure 13As shown, this is the tracking effect between the actual driving trajectory and the planned trajectory of the vehicle under actual conditions (with a vehicle motion model).

[0145] During the research and development process, the inventors discovered that when the control module performs algorithm processing, in order to ensure real-time performance and solution speed, it often uses small-angle assumptions for algorithm acceleration. However, in parking scenarios with large curvature changes or large steering angle changes, this introduces accumulated errors, severely impacting parking effectiveness and safety. For example... Figures 14-19 As shown, this illustrates the range in which the small-angle assumption of trigonometric functions is valid and the trend of error change after its failure.

[0146] exist Figure 14 , Figure 16 and Figure 18 In the diagram, the dark blue curve represents the true curve of the trigonometric function, while the red curve represents the approximate simplified formula curve of the corresponding trigonometric function.

[0147] exist Figure 15 , Figure 17 and Figure 19 In the diagram, the light blue curve represents the error between the true value of the trigonometric function and the simplified formula, the yellow line indicates the cutoff point when the error between the true value of the trigonometric function and the simplified formula is 1% (when the error between the simplified formula and the true formula is greater than 1%, the simplified formula will have an error when used), and the red circle indicates that the simplified formula can use the corresponding critical angle value.

[0148] When the closed-loop control accuracy is high (e.g., 1cm), the control law will be more sensitive. If the actual driving trajectory of the vehicle is more than 1cm to the left of the planned trajectory, in order to prevent the tracking error from increasing further, the controller will give the vehicle a right turn direction compensation. The same thing will happen when the vehicle turns right past the right side of the trajectory by 1cm. This repeated process will cause the steering wheel to vibrate back and forth when making large-angle turns.

[0149] In automatic parking scenarios, due to factors such as driving deviation, actuator wear, and positioning delay, the vehicle cannot strictly follow the planned trajectory. It is necessary to select a point to follow based on the planned trajectory. This point is called the forward view point. The forward view point is usually in front of the vehicle's driving direction and can be dynamically selected based on the current vehicle speed, the error between the vehicle and the planned trajectory, and the current driving scenario. The driving scenario can include straight lines, turns, high speeds, and low speeds. For example, the second tracking position is the forward view point before correction, and the first tracking position is the forward view point after correction.

[0150] In related technologies, a foreseeability point is typically selected on a pre-planned trajectory. However, the vehicle control method provided in this application does not select a foreseeability point on the planned trajectory, but rather selects a foreseeability point outside the planned trajectory points based on geometric constraints. Figures 20-22The image shows a comparison between the forward view selection correction of the vehicle control method provided in this application embodiment and the forward view processing in related technologies.

[0151] Among them, such as Figure 20 As shown, this illustrates the trajectory difference between the vehicle understeer and the planned trajectory; Figure 21 As shown, this illustrates a common method for handling the foreground point in related technologies, namely, selecting other foreground points along an existing planned trajectory; for example... Figure 22 The image shows a method for processing the forward viewpoint in the vehicle control method provided in this application embodiment, namely, correcting the position of the forward viewpoint based on computational geometry.

[0152] like Figure 23 The image shows a specific embodiment of the vehicle control method provided in this application.

[0153] Step 1: Perform a sharp turn and park.

[0154] Step 2: Collect wheel rotation angle, planned trajectory, forward viewpoint coordinates, and vehicle driving coordinates.

[0155] Step 3: Visualize the collected data offline.

[0156] Step 4: Determine whether the selected foresight point is on the planned trajectory.

[0157] Step 5: If the foreground point is not on the planned trajectory, it is assumed that a geometric model will be used to process the foreground point.

[0158] It should be noted that the forward viewpoint is dynamically changing. The vehicle control method provided in this application embodiment can correct the forward viewpoint based on a geometric model (wherein, the forward viewpoint may include a first tracking position and a second tracking position, the second tracking position is the initially selected forward viewpoint, and the first tracking position is the forward viewpoint obtained after correcting the second tracking position, for example, the second tracking position can be corrected through a geometric model to obtain the first tracking position). When selecting the forward viewpoint, it is not necessary to select it on an existing planned trajectory.

[0159] In the vehicle control method provided in this application embodiment, the forward viewpoint driving trajectory is adjusted in real time. The vehicle cannot reach the true forward viewpoint position, but it can improve the tracking effect of the vehicle's actual driving trajectory on the planned trajectory. In related technologies, the conventional forward viewpoint driving trajectory follows the planned trajectory, but when actually following, the vehicle's actual driving trajectory will have a large lateral error with the planned trajectory.

[0160] In actual implementation, the vehicle-mounted display device can be configured with a parking debugging mode. For example, when entering the vehicle model in the settings interface, the parking debugging model can be triggered by clicking the vehicle model icon multiple times, or by pressing a button or using voice, or by other means. This application does not limit this.

[0161] When the parking test mode is triggered, the rear vehicle display device can display real-time visual information from modules such as planning, control, radar, and vision sensors.

[0162] In this application, the global planned trajectory of the vehicle, the future driving trajectories of the four wheels of the vehicle, the forward viewpoint of the vehicle, the actual driving trajectory of the vehicle, the vehicle occupies a grid map, and a grid map of planar obstacles from a top-down perspective can be displayed on the display device.

[0163] In related technologies, a first- or second-order equation is usually obtained by fitting the vehicle speed and calibrated quantity to determine the forward viewpoint, or the problem is directly treated as a strongly nonlinear problem and solved iteratively using neural networks in deep learning.

[0164] The vehicle control method provided in this application embodiment establishes a constraint-based geometric model. By dynamically correcting the position of the forward viewpoint (second tracking position), the corrected forward viewpoint (first tracking position) is no longer on the original planned trajectory. By controlling the vehicle to turn using the corrected forward viewpoint, the accumulation of errors and singular perturbations of the control law during large-angle turns in automatic parking can be reduced, thereby improving the tracking performance during large-angle turns in parking.

[0165] The vehicle provided in this application embodiment can be any model, can be controlled by any implementable control algorithm, and the implementation process can be combined according to the actual situation. It has good universality and is less dependent on external information. It does not involve strong nonlinear models and high-dimensional matrix solutions, thus reducing the computational power consumption of the system.

[0166] like Figure 24 The image shows a specific embodiment of the vehicle control method provided in this application.

[0167] Step 1: Perform a sharp turn and park.

[0168] Step 2: Collect data on wheel angle, steering wheel angle, vehicle lateral tracking error, and vehicle coordinates. Step 3: Visualize the collected data offline.

[0169] Step 4: Analyze the vehicle steering tracking characteristics.

[0170] Step 5: Make a judgment based on the vehicle's lateral tracking error, steering wheel swing amplitude, and steering wheel swing frequency during large-angle steering.

[0171] Assuming the parking location is The closest point to the planned trajectory where the car is parked is the second tracking position. The next point after the nearest point on the trajectory is The car's position is C, and the forward viewpoint it needs to pass through to park is... The specific implementation process is as follows: Step 1: Determine if the shortest vertical distance from the vehicle's current position to the planned trajectory is less than a set value. Based on the lateral projection distance from the vehicle to the nearest point on the planned trajectory, determine whether the vehicle is in the parking process or in the parking preparation stage. If the following condition is met, it is considered to be in the parking preparation stage:

[0172] If it is the parking preparation stage, then it will be handled according to the second tracking position, where, It is a pre-set distance threshold. It is a vector pointing from the nearest point on the planned trajectory to the next point on the trajectory that is the nearest point to the vehicle. It is a vector pointing from the point closest to the vehicle on the planned trajectory to the vehicle's position.

[0173] Step 2: Based on the vehicle design parameters or actual calibration results, obtain the vehicle's accurate minimum turning radius (left / right), and fit the virtual wheel angle and front and rear wheelbase of the actual vehicle. The turning radius calibration can be performed offline and online based on the least squares method. The offline reference identification matrix is ​​as follows:

[0174] in, It is a general matrix expression for least squares identification, without any special physical meaning. Expanded into the world coordinates of multiple actual trajectory points collected. Axis coordinate queue, For a diagonal matrix that matches the dimension of matrix calculation, For collecting the world coordinates of multiple actual trajectory points Axis coordinate queue, The three coefficients are those of the formula to be identified. To collect noise and interference from real-world data when collecting vehicle trajectory points, To identify the corrected vehicle's true minimum turning radius.

[0175] If using online recursive least squares identification, then it is done by referring to the following formula:

[0176] The above formula is the online identification reasoning matrix formula for system identification, where the variables are all matrices representing abbreviations of intermediate quantities and have no special physical meaning.

[0177] like Figure 25 The figure shows the fitting effect of the vehicle's turning radius.

[0178] Step 3: Based on the results of Dead Reckoning (DR) or Trajectory Prediction (TP), determine whether the trajectory ahead of the current vehicle is a straight line, a small-angle turn, or a large-angle turn. If it's a straight line or a small-angle turn, no special processing is required; if it's a large-angle turn, the marker position is adjusted. Set to True.

[0179]

[0180] in, The first derivative of the trajectory fitting formula is given. The second derivative of the trajectory fitting formula is given by... To determine whether the current road is a sharp curve, a curvature threshold is used.

[0181] Step 4: After entering a sharp turn, establish a vehicle kinematic model based on the parking identification parameters. Calculate the driving difference between the ideal vehicle model and the actual vehicle during understeering (forward / reverse). Based on this, establish a constrained geometric model. Since this model is a convex set, it can be solved using analytical or optimization methods. Using analytical methods, the analytical solution can be obtained offline and directly written into the program, saving reliance on third-party libraries and reducing the computational burden during program execution. If the platform has sufficient computing power, existing optimization libraries can be directly called, treating it as a quadratic programming problem, with the solution expression being:

[0183] in, This is the sum of the cumulative lateral errors between the simulated trajectory and the planned trajectory. This is the maximum turning angle of the vehicle. To determine the path curvature for the planned trajectory, To simulate the path curvature of the trajectory, The origin of the vehicle coordinate system is the natural coordinate system. These are the coordinates of the road boundary in the natural coordinate system. For half the length of the car, Let be the origin of the vehicle's coordinates at any given time in the natural coordinate system. and These are the lower and upper bound conditions for preventing a vehicle from colliding with the road when it accelerates or decelerates. This represents the distance traveled by the vehicle in each unit of time within the natural coordinate system. These represent the lower bounds of the vehicle's boundary distance, velocity, acceleration, and jerk on the vehicle's coordinate axes in the natural coordinate system at any given time. These represent the boundary distance, velocity, acceleration, and upper bound of the jerk of the vehicle on the vehicle's coordinate axes in the natural coordinate system at any given time.

[0184] Step 5: Based on the feedback calculated by the constraint model, identify the vehicle's second tracking position and the first tracking position corrected by the geometric model. Then, based on the vector relationship between them, move the original second tracking position to the first tracking position to correct the difference between the vehicle's actual trajectory and the planned trajectory. After vector correction, the controller can calculate the control output for trajectory tracking to achieve the desired result. Figure 22 The effect shown.

[0185] like Figure 20 As shown, this illustrates the trajectory difference between the vehicle understeer and the planned trajectory. Figure 22 The image shows the corrected actual trajectory of the vehicle.

[0186] The vehicle control method provided in this application improves the tracking accuracy of large-angle steering during automatic parking by correcting the difference between the actual driving trajectory and the planned trajectory of the vehicle. Furthermore, through a geometric constraint model, the state-space equation based on the control law is always in a critical state of correction (ensuring tracking accuracy while keeping the output virtual steering angle of the front wheels at its maximum value), thus avoiding singular perturbations of the closed-loop controller when the tracking accuracy is high.

[0187] In the description of this application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc., indicating the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0188] In the description of this application, "first feature" and "second feature" may include one or more of the features.

[0189] In the description of this application, "multiple" means two or more.

[0190] In the description of this application, the first feature being "above" or "below" the second feature may include the first and second features being in direct contact, or the first and second features being in contact through another feature between them.

[0191] In the description of this application, the terms "above," "over," and "on top" for the first feature and the second feature include the first feature being directly above or diagonally above the second feature, or simply indicate that the first feature is at a higher horizontal level than the second feature.

[0192] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0193] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0194] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

Claims

1. A vehicle, characterized in that, When the vehicle travels to a turning section, between the starting position and the ending position of the turning section, the steering angle of the vehicle is controlled to change in the form of multiple peaks within a target frequency range, and the target frequency range is between 5 times per 5 seconds and 10 times per 5 seconds.

2. The vehicle according to claim 1, characterized in that, The vehicle's speed is any value between 3 km / h and 5 km / h, and the turning radius of the turning section is any value between 5 m and 8 m.

3. The vehicle according to claim 1, characterized in that, The turning section is a parking section. When the vehicle is parking and the predicted turning angle of the vehicle is greater than the preset angle, the steering angle of the vehicle changes in the form of multiple peaks within the target frequency range.

4. The vehicle according to any one of claims 1-3, characterized in that, The error between the vehicle's actual driving trajectory and the planned trajectory is within the target error range, which is between -15cm and 25cm. The actual driving trajectory is the trajectory obtained by the vehicle driving with the current driving parameters.

5. The vehicle according to claim 4, characterized in that, The target error range is between -10cm and 10cm between 5 seconds after the vehicle passes the starting position and 5 seconds before it passes the ending position.

6. The vehicle according to claim 4, characterized in that, The current driving parameters are determined based on the first tracking position. At any given moment, the first tracking position and the center of curvature of the planned trajectory are located on opposite sides of the planned trajectory.

7. The vehicle according to claim 6, characterized in that, The vehicle includes a rear axle, and in a top-down view, the straight-line distance between the center of the rear axle and the first tracking position satisfies the target spacing range.

8. The vehicle according to claim 7, characterized in that, The target spacing ranges from 2m to 12m.

9. The vehicle according to claim 6, characterized in that, The vehicle includes a control module for determining a first tracking position. The first tracking position is determined based on a second tracking position located on the planned trajectory. The first tracking position is located on a simulated trajectory, and the simulated trajectory is the trajectory corresponding to the vehicle's simulated driving with the current driving parameters.

10. The vehicle according to claim 9, characterized in that, With the goal of minimizing the deviation between the simulated trajectory and the planned trajectory, and while satisfying the vehicle's driving constraints, the first tracking position is determined.

11. A vehicle, characterized in that, When the vehicle is traveling on a curve, the current driving parameters are determined based on the first tracking position. At any given moment, the first tracking position and the center of curvature of the planned trajectory of the vehicle are located on opposite sides of the planned trajectory.

12. The vehicle according to claim 11, characterized in that, The vehicle includes a rear axle, and in a top-down view, the straight-line distance between the center of the rear axle and the first tracking position satisfies the target spacing range.

13. The vehicle according to claim 11 or 12, characterized in that, The vehicle includes a control module for determining a first tracking position. The first tracking position is determined based on a second tracking position located on the planned trajectory. The first tracking position is located on a simulated trajectory, and the simulated trajectory is the trajectory corresponding to the vehicle's simulated driving with the current driving parameters.

14. The vehicle according to claim 13, characterized in that, With the goal of minimizing the deviation between the simulated trajectory and the planned trajectory, and while satisfying the vehicle's driving constraints, the first tracking position is determined.

15. The vehicle according to claim 14, characterized in that, The driving constraints include at least one of the following: the simulated trajectory is within the target driving range, the vehicle's speed is within the target speed range, the vehicle's acceleration is within the target acceleration range, and the curvature of the vehicle's simulated trajectory is within the target curvature range.

Images