Vehicle obstacle avoidance implementation method, system, vehicle and storage medium

Abstract

This application provides a vehicle obstacle avoidance method, system, vehicle, and storage medium. The method includes: responding to an obstacle avoidance warning command, calculating the collision point position based on the current preset center position of the vehicle and the target longitudinal distance xend for obstacle avoidance; obtaining the target lateral distance yend for obstacle avoidance based on the collision point position and a preset threshold vehicle lateral offset ratio to obtain the target obstacle avoidance starting point; calculating the target obstacle avoidance parameters based on the current vehicle speed, or the current vehicle speed and the current speed of the obstacle, and outputting an obstacle avoidance steering command only when the current actual obstacle avoidance parameters of the vehicle meet the target obstacle avoidance parameters; and responding to the obstacle avoidance steering command, performing longitudinal and lateral sampling accordingly to generate an obstacle avoidance path plan. This application, by comprehensively considering longitudinal and lateral constraints as well as obstacle constraints, accurately calculates the obstacle avoidance starting point and path ending point under extreme obstacle avoidance conditions, as well as the precise sampling path, which can significantly improve the obstacle avoidance success rate.

Description

Technical Field

[0001] This application relates to the field of vehicle obstacle avoidance technology, and in particular to a method, system, vehicle, and storage medium for implementing vehicle obstacle avoidance. Background Art

[0002] In autonomous driving, emergency steering and obstacle avoidance can be implemented when a vehicle approaches a target obstacle at high speed. If emergency braking cannot prevent a collision, the system determines whether steering is an option for emergency obstacle avoidance. If steering is deemed feasible, the system plans an appropriate obstacle avoidance path for the vehicle to execute. However, existing technologies lack theoretical calculations to accurately calculate the closest obstacle avoidance path to the obstacle under specific conditions of acceleration, curvature, and obstacle boundary. Current technologies can only estimate the latest point at which obstacle avoidance can be activated but cannot precisely determine the corresponding value. Even when offline calibration methods are used, these are only rough judgments based on experimental data and cannot perform precise calculations. Furthermore, due to the inability to calculate precisely, existing technologies primarily rely on sampling to determine passable obstacle avoidance paths; therefore, large-scale, high-granularity sampling results in a massive computational burden. Since the characteristics of the obstacle avoidance path under extreme conditions of lateral acceleration or curvature cannot be determined, sampling needs to be carried out within the range of the starting and ending points of the lateral and longitudinal sampling. Although some algorithms in the existing technology have optimized the sampling points, the constraints of the obstacle avoidance path cannot be accurately calculated. Excessive sampling intervals will cause the vehicle to miss the original possible obstacle avoidance path. Although the operating load of the controller is reduced, excessive sampling intervals may significantly reduce the effective range of the function, thereby causing collisions in scenes that could have avoided obstacles, resulting in personal injury or financial loss. Summary of the Invention

[0003] To address the shortcomings of the existing technologies, this application aims to provide a vehicle obstacle avoidance method, system, vehicle, and storage medium. By comprehensively considering lateral and longitudinal constraints as well as obstacle constraints, the method accurately calculates the obstacle avoidance start point and path end point under extreme obstacle avoidance conditions. Furthermore, by accurately sampling the lateral and longitudinal target distances for obstacle avoidance steering based on the precisely calculated distances, the method can effectively improve the sampling success rate, reduce the number of samples, improve execution efficiency, and significantly increase the obstacle avoidance success rate.

[0004] The vehicle obstacle avoidance method and system proposed in this application can be widely applied in various driving scenarios, such as: autonomous vehicles, where the method and system can be applied to obstacle avoidance decision-making and path planning to improve the safety and efficiency of obstacle avoidance; driver assistance systems, where the method can be integrated into active safety systems such as automatic emergency braking and automatic steering to assist drivers in obstacle avoidance; intelligent transportation systems, where it can be applied to public transportation systems such as buses and trucks to assist or replace drivers in obstacle avoidance operations; intelligent logistics vehicles, where it can be applied to unmanned forklifts and handling vehicles in warehousing and logistics to achieve their automatic obstacle avoidance capabilities; intelligent robots, where it can be applied to various intelligent robots that require autonomous obstacle avoidance capabilities, such as service robots and exploration robots; and intelligent toys, where it can be applied to intelligent toy cars and robots that require obstacle avoidance functions.

[0005] In a first aspect, the present invention provides a vehicle obstacle avoidance method, wherein the direction in which the vehicle encounters an obstacle during driving is defined as the longitudinal direction X, and the direction perpendicular to the driving road surface is defined as the transverse direction Y, comprising:

[0006] S1: In response to the obstacle avoidance warning command, based on the current preset center position of the vehicle and the longitudinal distance x of the target obstacle to be avoided by the vehicle. end This allows us to calculate the location of the collision point on one side of the vehicle.

[0007] S2: Based on the collision point location on one side of the vehicle and a preset threshold vehicle lateral offset ratio, obtain the target lateral distance y for obstacle avoidance by the vehicle. end To obtain the starting point for obstacle avoidance.

[0008] S3: Calculate the target obstacle avoidance parameters based on the vehicle's current speed, or the vehicle's current speed and the obstacle's current speed, and output the obstacle avoidance steering command only when the vehicle's current actual obstacle avoidance parameters meet the target obstacle avoidance parameters.

[0009] S4: In response to the obstacle avoidance steering command, based on the target longitudinal distance x end Vertical sampling is performed according to preset rule one, and based on the target lateral distance y end Lateral sampling is performed according to preset rule two to generate obstacle avoidance path planning.

[0010] In this application, prior to step S1, the following steps are also included:

[0011] Obtain the curvature value of the vehicle traveling on the current path and the vehicle's current lateral controllable threshold acceleration.

[0012] Based on the curvature value and the lateral controllable threshold acceleration, the target longitudinal distance x for vehicle obstacle avoidance is calculated. end .

[0013] In this application, step S1 specifically includes:

[0014] When the obstacle avoidance warning command is received, the longitudinal lane change completion rate when the vehicle collides with the obstacle is obtained according to the preset threshold vehicle lateral offset ratio.

[0015] Then, based on the target longitudinal distance x end Based on the longitudinal lane change completion rate, the location of the collision point on one side of the vehicle is calculated.

[0016] In this application, step S2 specifically includes:

[0017] The target lateral distance y for vehicle obstacle avoidance end This includes the lateral distance y of multiple targets.

[0018] Based on the location of the collision point on one side of the vehicle and the proportion of the vehicle's lateral offset that varies at preset intervals within a preset threshold range, multiple sets of target lateral distances y for vehicle obstacle avoidance are calculated.

[0019] Based on the vehicle speed values ​​at preset intervals and the target lateral distance y for vehicle obstacle avoidance, multiple sets of target obstacle avoidance starting points with vehicle lateral offset ratios varying at preset intervals are obtained.

[0020] In this application, the target obstacle avoidance parameters include the target obstacle avoidance distance.

[0021] The target obstacle avoidance distance is calculated based on the vehicle's current speed, the obstacle's current speed, and the preset collision compensation distance.

[0022] Furthermore, the target obstacle avoidance parameters also include the target obstacle avoidance time.

[0023] The target obstacle avoidance time is calculated based on the vehicle's current speed and the preset collision compensation time.

[0024] The obstacle avoidance steering command is output only when the vehicle's current actual obstacle avoidance distance is greater than the target obstacle avoidance distance and the vehicle's current actual obstacle avoidance time is greater than the target obstacle avoidance time.

[0025] In this application, the step of determining the target longitudinal distance x end Vertical sampling is performed according to preset rule one, specifically as follows:

[0026] With the target longitudinal distance x end Based on the first baseline, longitudinal sampling is performed according to the preset back-off and preset forward sampling intervals.

[0027] The target lateral distance y end Lateral sampling is performed according to preset rule two, specifically as follows:

[0028] With the target lateral distance y end Based on the second reference, lateral sampling is performed according to a preset lateral sampling interval; wherein, when the lateral sampling point exceeds the lane boundary range, the boundary value is taken as the lateral sampling point.

[0029] In a second aspect, the present invention also provides a vehicle obstacle avoidance system, comprising: a first processing unit, a second processing unit, a third processing unit, and a fourth processing unit.

[0030] The first processing unit is used to respond to obstacle avoidance warning commands and, based on the current preset center position of the vehicle and the longitudinal distance x of the target obstacle to be avoided, performs the following operations. end This allows us to calculate the location of the collision point on one side of the vehicle.

[0031] The second processing unit is used to obtain the target lateral distance y for obstacle avoidance of the vehicle based on the collision point location on one side of the vehicle and a preset lateral offset ratio of the vehicle. end To obtain the starting point for obstacle avoidance.

[0032] The third processing unit is used to calculate the target obstacle avoidance parameters based on the vehicle's current speed, or the vehicle's current speed and the obstacle's current speed, and to output an obstacle avoidance steering command only when the vehicle's current actual obstacle avoidance parameters meet the target obstacle avoidance parameters.

[0033] The fourth processing unit is configured to respond to the obstacle avoidance steering command and, based on the target longitudinal distance x, end Vertical sampling is performed according to preset rule one, and based on the target lateral distance y end Lateral sampling is performed according to preset rule two to generate obstacle avoidance path planning.

[0034] In a third aspect, the present invention also provides a vehicle that includes at least the vehicle obstacle avoidance system described above.

[0035] In a fourth aspect, the present invention also provides a storage medium, which is a type of computer-readable storage medium, having stored thereon a computer program that, when executed by a processor, implements the vehicle obstacle avoidance method as described above.

[0036] Compared with the prior art, this application has at least the following beneficial effects:

[0037] This application proposes a vehicle obstacle avoidance method, system, vehicle, and storage medium. By comprehensively considering lateral and longitudinal constraints as well as obstacle limitations, it accurately calculates the obstacle avoidance start point and path endpoint under extreme obstacle avoidance conditions. It can cover various working scenarios and significantly improves the obstacle avoidance success rate. This application also accurately calculates the target collision time of the minimum obstacle avoidance turn or the target collision distance of the latest obstacle avoidance turn, which is used by the system to determine whether to execute the obstacle avoidance turn function, the trajectory of vehicle instability, and the risk of collision. Based on the accurately calculated lateral and longitudinal target distances for obstacle avoidance turn, this application performs lateral and longitudinal sampling, which effectively improves the sampling success rate, reduces the number of samples, and improves execution efficiency. Furthermore, it accurately judges the validity of trajectory constraints on the generated sampled paths, which can significantly improve code execution efficiency. This application greatly reduces the risk of collisions in vehicle obstacle avoidance scenarios, largely avoiding personal injury or financial loss. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating a vehicle obstacle avoidance method according to an embodiment of this application.

[0039] Figure 2 This is a schematic diagram illustrating an obstacle avoidance method for vehicles according to an embodiment of this application.

[0040] Figure 3 The collision distance x at different vehicle speeds in the embodiments of this application. a-max The offline calculation results of the corresponding vehicle lateral offset ratio are shown in the figure.

[0041] Figure 4 This is a structural block diagram of a vehicle obstacle avoidance system according to an embodiment of this application. DETAILED DESCRIPTION

[0042] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0043] Example 1:

[0044] Please refer to the appendix. Figure 1 , Figure 1This is a flowchart illustrating a vehicle obstacle avoidance method provided in an embodiment of this application. The vehicle obstacle avoidance method of this application can be widely applied in various driving scenarios, such as: autonomous vehicles, where the method and system can be applied to obstacle avoidance decision-making and path planning to improve the safety and efficiency of obstacle avoidance; driver assistance systems, where the method can be integrated into active safety systems such as automatic emergency braking and automatic steering systems to assist the driver in obstacle avoidance; intelligent transportation systems, which can be applied to public transportation systems such as buses and trucks to assist or replace drivers in obstacle avoidance operations; intelligent logistics vehicles, which can be applied to unmanned forklifts and handling vehicles in warehousing and logistics to achieve their automatic obstacle avoidance capabilities; intelligent robots, which can be applied to various intelligent robots requiring autonomous obstacle avoidance capabilities, such as service robots and exploration robots; and intelligent toys, which can be applied to intelligent toy cars and robots requiring obstacle avoidance functions, and are not limited to these.

[0045] As attached Figure 2 The diagram illustrates a specific scenario analysis of a vehicle obstacle avoidance method in this embodiment. This method uses the direction in which the vehicle encounters an obstacle as the longitudinal direction (X), and the direction perpendicular to the road surface as the lateral direction (Y). The following analysis is based on the aforementioned reference directions. The method mainly includes:

[0046] S1: In response to the obstacle avoidance warning command, based on the current preset center position of the vehicle and the longitudinal distance x of the target obstacle to be avoided by the vehicle. end This allows us to calculate the location of the collision point on one side of the vehicle.

[0047] It should be noted that when a vehicle is traveling at a certain speed in a lane and an obstacle appears in front of it, once the obstacle enters the detectable range, it will be detected by the vehicle's sensors and other detection modules, and an obstacle avoidance warning command signal will be sent to the vehicle control terminal. The vehicle control terminal will respond to the obstacle avoidance warning command and obtain the relative distance between the vehicle and the obstacle. This is because the obstacle may be a stationary object or a vehicle that has suddenly decelerated or stopped in the lane due to other unexpected road malfunctions. The vehicle control terminal of this application will collect the vehicle's current driving parameters and road conditions in real time, such as vehicle acceleration, lane curvature, etc., and accurately calculate the target longitudinal distance x for obstacle avoidance based on these parameters. end Wherein, the target's longitudinal distance x end It refers to the shortest vertical distance x. endThis refers to the shortest longitudinal path for the vehicle to change lanes and avoid obstacles, provided that all surrounding environmental parameters are permissible. Additionally, the vehicle's preset center position can be the rear axle center. Combined with pre-calculated longitudinal lane-change completion rates based on corresponding vehicle speed, vehicle width, and different threshold lateral offset ratios, the collision point location on one side of the vehicle can be calculated more precisely. Based on this accurate collision point location, collision risk can be assessed, providing a basis for determining the timing of obstacle avoidance. Furthermore, it provides key parameters for calculating the optimal obstacle avoidance starting point, planning the optimal obstacle avoidance path, and achieving precise obstacle avoidance steering time control, thus avoiding dangerous obstacle avoidance maneuvers and improving the overall performance of the obstacle avoidance system.

[0048] S2: Based on the collision point location on one side of the vehicle and a preset threshold vehicle lateral offset ratio, obtain the target lateral distance y for obstacle avoidance by the vehicle. end To obtain the starting point for obstacle avoidance.

[0049] It should be noted that after calculating the collision point on one side of the vehicle, it is necessary to determine the lateral distance y of the obstacle avoidance target. end To obtain the optimal obstacle avoidance starting point. The optimal obstacle avoidance starting point refers to the latest point, after analysis and calculation, that allows successful obstacle avoidance and enables the activation of the obstacle avoidance function, ensuring that the obstacle avoidance method of this application can be successfully implemented even under the most extreme conditions. Target lateral distance y end This refers to the shortest lateral offset distance from the vehicle's centerline to the obstacle avoidance endpoint, which determines the obstacle avoidance trajectory. Different ranges and intervals for different vehicle lateral offset ratios are pre-set based on experience. Based on the collision point location on one side of the vehicle and different preset vehicle lateral offset ratios, multiple candidate values ​​of y can be calculated. Preferably, based on a preset vehicle speed range, the obstacle avoidance trajectory under each set of y is calculated, and the optimal y is determined. end Make the obstacle avoidance path the shortest.

[0050] This allows us to obtain the target's lateral distance y that minimizes the obstacle avoidance distance under the current conditions. end y end Combined with x end The target starting point for obstacle avoidance can be determined, making the obstacle avoidance strategy more precise. The optimal value for y is found by varying the vehicle's lateral offset ratio. end This achieves optimized selection of obstacle avoidance starting point.

[0051] S3: Calculate the target obstacle avoidance parameters based on the vehicle's current speed, or the vehicle's current speed and the obstacle's current speed, and output the obstacle avoidance steering command only when the vehicle's current actual obstacle avoidance parameters meet the target obstacle avoidance parameters.

[0052] It should be noted that when there is relative motion between the obstacle and the current vehicle, obstacle avoidance must consider not only the vehicle's own state but also the obstacle's motion state. Target obstacle avoidance parameters include target avoidance distance and target avoidance time. The target avoidance distance is calculated based on the vehicle speed, obstacle speed, and a preset collision safety distance; the target avoidance time is calculated based on the vehicle speed and a preset collision safety time. The target avoidance parameters are compared with the current actual obstacle avoidance state. An obstacle avoidance steering command is only output when both the actual avoidance distance and time are greater than the target parameters. If the actual distance or time is less than or equal to the target parameters, there is a risk of collision, and obstacle avoidance is not performed.

[0053] This allows for the determination of the latest safe obstacle avoidance opportunity, preventing dangerous obstacle avoidance maneuvers. Considering kinematic constraints and providing safety redundancy enhances obstacle avoidance safety.

[0054] S4: In response to the obstacle avoidance steering command, based on the target longitudinal distance x end Vertical sampling is performed according to preset rule one, and based on the target lateral distance y end Lateral sampling is performed according to preset rule two to generate obstacle avoidance path planning.

[0055] It should be noted that after outputting the obstacle avoidance steering command, a specific obstacle avoidance path needs to be generated. Path generation uses a sampling method, sampling multiple points both longitudinally and laterally. The longitudinal sampling uses a pre-calculated target distance x. end Sampling is performed according to preset rules based on a baseline; lateral sampling can be performed using a pre-calculated y end Based on this, samples are also taken according to preset rules. Finally, curve fitting is performed between the sampled points to generate the final obstacle avoidance path.

[0056] This sampling strategy can improve computational efficiency and avoid a large number of invalid samples. By using pre-calculated key parameters for targeted sampling, the path becomes more feasible and optimized, significantly reducing the computational load during the sampling process.

[0057] In this application, prior to step S1, the following steps are also included:

[0058] Obtain the curvature value of the vehicle traveling on the current path and the vehicle's current lateral controllable threshold acceleration.

[0059] Based on the curvature value and the lateral controllable threshold acceleration, the target longitudinal distance x for vehicle obstacle avoidance is calculated. end .

[0060] Preferably, the obstacle avoidance planning trajectory calculation in this application is performed using a fifth-order polynomial. Based on the characteristics of a fifth-order polynomial, the parameters for the starting and ending points are obtained as shown in Equations 1 and 2 below:

[0061] y start =0y′ start =0y'' start =0,

[0062] y end =y end y′ end =0y'' end =0.

[0063] Based on the above formula, we can obtain the simplified form of the fifth-degree polynomial, Equation 3:

[0064]

[0065] Differentiating the above equation, we obtain equation 4:

[0066]

[0067] Then, by differentiating again, we obtain Equation 5:

[0068]

[0069] Where, x end The target longitudinal distance is required to be obtained from the endpoint of the fifth-order polynomial, and x / x end This indicates the completion rate of the longitudinal lane change.

[0070] Furthermore, Equation 4 above represents the tangent of the heading angle θ at each point on the trajectory. Considering that θ is very small during the lane change process, tan(θ)≈θ, which leads to Equation 6:

[0071]

[0072] Meanwhile, according to the lane curvature formula 7:

[0073]

[0074] It can be seen that, under actual road conditions, since θ is very small, it is assumed that... Then we can obtain equation 8:

[0075]

[0076] Taking the derivative of equation 8 above, when the derivative of k is 0, we obtain equation 9:

[0077]

[0078] Solving equation 9 above, we find that the extreme value of the curvature k of the fifth-degree polynomial occurs at x / x. end When x / x = 0.21 or 0.79, endSubstituting 0.21 or 0.79 into equation 5 simplifies to equation 10:

[0079]

[0080] Then combine the calculation of lateral acceleration a y Equation 11: Equation 12 can be obtained:

[0081] After adjusting and transforming Equation 12, we can obtain Equation 13:

[0082] Preferably, a ymax The value can be 3 m / s², or it can be determined by the total acceleration limit a. max Subtracting longitudinal acceleration a x The calculation method is as follows: (Formula 14)

[0083]

[0084] Preferably, the total acceleration limit value a is at this time. max We can take 5 m / s². From this, we can calculate the shortest longitudinal distance x of the obstacle avoidance path under the critical condition of lateral acceleration when the lane curvature is 0, specifically on a straight road. end .

[0085] In another preferred embodiment, when considering curves, the lane centerline is used as a reference line. The vehicle's trajectory is generated using a fifth-order polynomial in the Frenet coordinate system, with the reference line as the reference. In the Frenet coordinate system, the longitudinal distance traveled along the reference line is denoted as S, and the lateral displacement relative to the reference line is denoted as l. The trajectory generated in the Frenet coordinate system considers not only the curvature in the Frenet coordinate system but also the curvature value inherent in the vehicle's current movement.

[0086] Equation 15 is obtained by simplifying the curvature calculation formula of the Frenet coordinate system: Where, k x Let k be the curvature in Cartesian coordinates. r For the lane curvature, considering k r It is very small, and k r The rate of change is very small.

[0087] Combining and simplifying equations 15 and 13, we obtain equation 16:

[0088]

[0089] It should be noted that, as shown in Equation 16, in a curve scenario, the lateral acceleration can be subtracted from the acceleration component caused by the road curvature at the current vehicle speed, and the remaining part can be substituted into the fifth-order polynomial maximum curvature calculation formula to calculate the shortest longitudinal distance x of the fifth-order polynomial. end .

[0090] Using the method described above, the shortest longitudinal distance x that satisfies the lateral acceleration constraint under curves of different curvatures can be calculated quickly. end This solves the technical problem in existing technologies that cannot accurately calculate the obstacle avoidance path closest to the obstacle under conditions of acceleration, curvature, and obstacle boundary.

[0091] In this application, step S1 specifically includes:

[0092] When the obstacle avoidance warning command is received, the vehicle's lateral offset ratio y is determined according to the preset threshold. ratio Obtain the longitudinal lane change completion rate x / x when the vehicle collides with the obstacle. end .

[0093] Then, based on the target longitudinal distance x end and the longitudinal lane change completion rate x / x end The location of the collision point on one side of the vehicle was calculated.

[0094] It should be noted that different y end It will affect x at the same time end The calculation and lane change completion rate are considered, so it is possible to calculate the collision distance x under different combinations of vehicle lateral offset ratio and vehicle speed during offline calculation. a_max The table is created, and in actual use, the y-value is derived by taking the vehicle's lateral offset ratio corresponding to the minimum collision distance at the current vehicle speed. end That is, in each round of offline calculation, a vehicle lateral offset ratio is first set to calculate y. end Then calculate the lane change completion rate and x. end Finally, the collision distance is calculated.

[0095] It should be noted that, in this application, based on the target lateral distance y end The lateral coordinate y of the point of collision with the obstacle on one side c Based on the vehicle width B, the lateral offset ratio y of the vehicle when it collides with the obstacle is calculated. ratio =(2y c +B) / 2y end Then adjust the vehicle's lateral offset ratio y. ratio Substituting into a fifth-order polynomial, we obtain the longitudinal lane change completion rate x. c / x endThen, based on the target longitudinal distance x end And the longitudinal lane change completion rate, calculate the distance x from the vehicle's preset center position to the collision point. c .

[0096] Preferably, Equations 17 and 18 can be obtained from the vehicle geometry:

[0097]

[0098]

[0099] Where (x) a ,y a Let B be any point on the side of the vehicle where a potential collision is possible, and let L be the width of the vehicle. a For (x) a ,y a The distance is the longitudinal distance from the origin (x, y) of the vehicle's coordinate system. Considering that θ will not be very large during the lane change, sin(θ) ≈ θ, and cos(θ) ≈ 1. Therefore, equations 17 and 18 can be simplified to equations 19 and 20 respectively:

[0100]

[0101]

[0102] Furthermore, substituting equations 3 and 4 one-to-one into equations 19 and 20 respectively, we obtain equations 21 and 22 as follows:

[0103]

[0104]

[0105] Furthermore, by simplifying and combining like terms in equations 21 and 22, equation 21 remains unchanged, while equation 22 is transformed to obtain equation 23 below:

[0106]

[0107] Preferably, it is assumed that when a collision occurs at point a, i.e., y a =y c , x a =x c Let x / x end =U, change y a =y c Substituting into equation 23, we obtain equation 24:

[0108]

[0109] It should be noted that solving Equation 24 above yields the longitudinal lane-change completion degree U when a collision occurs at point a. After calculating U, this degree is then combined with the x obtained from the calculation of the shortest lane-change path. end This allows us to calculate the distance traveled by the vehicle's center position A(x,y) relative to the starting point A(x0,y0) when the collision occurs at point a, i.e., x = U*x end Furthermore, the value of y can also be calculated using equations 19 and 20.

[0110] It should be noted that, considering the collision occurred at point a, the point of impact on one side of the vehicle at this time is (x). a ,y a The distance from the obstacle avoidance starting point (x0, y0) can also be calculated using Equation 19. In Equation 24, y c The collision margin can be calculated by adding a certain amount of collision margin to the width of the obstacle. Preferably, the collision margin is 0.2m.

[0111] Considering the inherent errors in sensor detection and control system calculations, relying solely on the actual width of the obstacle could lead to misjudgments. Therefore, increasing the collision margin provides tolerance and avoids misjudgments. This also enhances obstacle avoidance safety, as both the vehicle and the obstacle may exhibit some movement error during actual obstacle avoidance. Increasing the collision margin allows the vehicle to begin obstacle avoidance from a safe distance, reducing the risk of an actual collision. The margin allows the vehicle to avoid obstacles at a certain distance, reducing the possibility of the vehicle being "grazed" by the obstacle and preventing secondary collisions.

[0112] In this application, step S2 specifically includes:

[0113] The target lateral distance y for vehicle obstacle avoidance end This includes the lateral distance y of multiple targets.

[0114] Based on the location of the collision point on one side of the vehicle and the proportion of the vehicle's lateral offset that varies at preset intervals within a preset threshold range, multiple sets of target lateral distances y for vehicle obstacle avoidance are calculated.

[0115] Based on the vehicle speed values ​​at preset intervals and the target lateral distance y for vehicle obstacle avoidance, multiple sets of target obstacle avoidance starting points with vehicle lateral offset ratios varying at preset intervals are obtained.

[0116] It should be noted that in existing technologies, y is often used by default. end The value is taken as the center point of the adjacent lane in the obstacle avoidance direction. This method may not necessarily guarantee the shortest obstacle avoidance starting point, and when the obstacle in front intrudes into the adjacent lane in the obstacle avoidance direction, it may cause the obstacle avoidance path planning to fail.

[0117] However, y endIt's not necessarily true that the smaller the better, although due to y end Decrease will cause x end Shortened, but too small y end This will also increase the value of U during obstacle avoidance, which may lead to an increase in x. Therefore, it is necessary to find the shortest y-axis for obstacle avoidance. end .

[0118] This application achieves the selection of the optimal obstacle avoidance starting point through traversing and optimizing multiple sets of parameters. Among them, the target lateral distance y... end This includes multiple candidate solutions, denoted as y. Multiple sets of y are calculated based on a pre-set vehicle lateral offset ratio range and variation interval. The vehicle lateral offset ratio range is preset to 0.4-1, and the interval can be 0.1.

[0119] Preferably, the trajectory for each set of y is calculated according to a preset speed range and interval. The speed range can be preset to 40-100 km / h, and the interval can be 5 km / h. The trajectory length for each set of y is calculated at all speeds. The set of y with the shortest trajectory is selected as the final y. end That is, at various vehicle speeds, optimize and select the shortest obstacle avoidance trajectory corresponding to y. Finally, based on the optimal y... end The target starting point for obstacle avoidance is determined by the corresponding vehicle speed, but is not limited to this.

[0120] It should be noted that this application can solve for y under the shortest obstacle avoidance starting point using either online or offline methods. end value.

[0121] Preferably, the principle of the online real-time solution method is as follows:

[0122] First, obtain the vehicle's width B, the distance Lf from the coordinate center to the front face, and the distance -Lr from the coordinate center to the rear face. Then, obtain the lateral coordinate yc of the collision point on one side of the obstacle.

[0123] set up: And it iterates through each vehicle lateral offset ratio y at preset intervals, from 0.4 to 1, preferably at intervals of 0.1. ratio For each group of y ratio The corresponding y can be calculated. end Then, combining the current vehicle speed and the maximum lateral acceleration limit, x is calculated. end At regular intervals, such as 0.2m, traverse from -Lr to Lf, find and record the calculated x value for the collision point on one side of each vehicle. a The maximum value x a_max .

[0124] Then calculate y for each group. ratio All x under a_max Find x a_max The minimum value corresponding to yratio_min Then, based on y... ratio_min Calculate the minimum lane change endpoint y end_min .

[0125] The above-mentioned real-time online solution will be more accurate and safer and more reliable. However, considering that it requires solving Equation 24 multiple times, the amount of calculation is large and the time is long. It may not be suitable for some autonomous driving controllers with lower computing power in the short term when obstacle avoidance is achieved.

[0126] Preferably, this application can also be solved offline, as follows:

[0127] First, determine the vehicle's width B, the distance Lf from the coordinate center to the front face, and the distance -Lr from the coordinate center to the rear face. Then, determine the lateral coordinate y of the collision point on one side of the obstacle. c .

[0128] set up: And it iterates through each vehicle lateral offset ratio y at preset intervals, typically from 0.4 to 1, with intervals of 0.1. ratio For each group of y ratio The corresponding y can be calculated. end .

[0129] Next, at certain intervals, such as 5 km / h, the vehicle speed is iterated from 40 km / h to 100 km / h, and then x is calculated by combining the current vehicle speed and the maximum lateral acceleration limit. end Simultaneously, at regular intervals, such as 0.2m, traverse from -Lr to Lf to find and record the calculated x value at the collision point on one side of each vehicle. a The maximum value x a_max .

[0130] As attached Figure 3 As shown, based on the calculation results obtained above, a graph is constructed with the vehicle speed on the horizontal axis and the y-axis on the vertical axis. ratio The calibration table, each grid point is filled with the corresponding x a_max This is used for real-time table lookup during function operation, which allows the collision distance x at different vehicle speeds to be obtained. a-max The offline calculation results of the corresponding vehicle lateral offset ratio are shown in the figure.

[0131] Preferably, after the vehicle responds to the obstacle avoidance warning command while driving in the lane, it waits until the shortest obstacle avoidance starting point y is reached. end When calculating the value, first determine y based on the boundaries of adjacent obstacle avoidance lanes and surrounding obstacles. end_max Then calculate the corresponding y ratio_min And then from 1 to y ratio_min Within the range, and in conjunction with the vehicle's speed V, find x in the table above. a-max and the corresponding y ratioThe values ​​are not limited to these.

[0132] In this application, the target obstacle avoidance parameters include the target obstacle avoidance distance.

[0133] The target obstacle avoidance distance is calculated based on the vehicle's current speed, the obstacle's current speed, and the preset collision compensation distance.

[0134] Furthermore, the target obstacle avoidance parameters also include the target obstacle avoidance time.

[0135] The target obstacle avoidance time is calculated based on the vehicle's current speed and the preset collision compensation time.

[0136] The obstacle avoidance steering command is output only when the vehicle's current actual obstacle avoidance distance is greater than the target obstacle avoidance distance and the vehicle's current actual obstacle avoidance time is greater than the target obstacle avoidance time.

[0137] It should be noted that the target obstacle avoidance parameters include the target obstacle avoidance distance and the target obstacle avoidance time. The target obstacle avoidance distance refers to the minimum obstacle avoidance distance (DIST), and the target obstacle avoidance time refers to the latest obstacle avoidance time (TTC). Based on the above calculations, the closest starting point for obstacle avoidance during steering (x0, y0) from the rear axle center to the point of impact on one side of the vehicle can be obtained. 0) However, since there may be relative motion between the obstacle and the vehicle, the closest starting point for this steering obstacle avoidance (x0, y0) is... 0) It also needs to be converted into the latest obstacle avoidance distance or the corresponding minimum TTC when there is a certain relative motion between the obstacle and the vehicle. min value.

[0138] This application requires simultaneous calculation of the latest obstacle avoidance time (TTC). min The values ​​and minimum obstacle avoidance distance (DIST) are used to address different system requirements.

[0139] Preferably, the calculation and analysis of the minimum obstacle avoidance distance (DIST) value is shown in Equation 25 below:

[0140]

[0141] Among them, V obs Where is the absolute speed of the obstacle, and Dist0 is the collision safety distance calculated by comprehensively considering driver reaction time, system response delay, and collision safety margin. The current relative distance between the vehicle and the obstacle is Dist > DIST. min When Dist is not greater than DIST, the first obstacle avoidance steering function is allowed to be triggered. min At that time, there is a risk of collision, and the system will not allow the obstacle avoidance steering function to be triggered.

[0142] Preferably, the latest obstacle avoidance time (TTC) min The calculation and analysis of the value are shown in Equation 26 below:

[0143]

[0144] Among them, V ego Let T0 be the absolute speed of the vehicle, and T0 be the compensation time calculated by comprehensively considering driver reaction time, system response delay, and safety margin. Considering x... a It is a longitudinal distance calculated based on the vehicle's center (x0, y0) as the origin. When calculating the collision risk, the collision point on one side of the vehicle (x0, y0) needs to be subtracted. a ,y a The distance between the vehicle's center (x0, y0) and the vehicle's center can be used to calculate the TTC (Total Traction Cost) after subtracting the entire length of Lf for safety reasons, although a potential side collision could occur. min When a potential collision point on one side of the vehicle occurs at another point less than Lf distance away, the aforementioned TTC provides a certain time margin and is relatively safer. min This can be used as the basis for determining the latest obstacle avoidance turning time, that is, when the actual TTC calculated based on the vehicle's speed and relative distance is not greater than the TTC. min When TTC > TTC, no steering obstacle avoidance maneuver is performed; only when TTC > TTC. min The second obstacle avoidance steering function is only allowed to be triggered when other steering and obstacle avoidance conditions are met.

[0145] It should be noted that the onboard autonomous driving controller will only execute obstacle avoidance steering and proceed to the next path sampling mode when both the first and second obstacle avoidance steering functions are triggered simultaneously. This dual-module simultaneous judgment mechanism can more reliably and accurately determine the optimal obstacle avoidance timing, improving the effectiveness of obstacle avoidance while ensuring safety. This ensures obstacle avoidance can be achieved even in extreme situations, improving the safety of obstacle avoidance judgment. The dual condition constraints not only reduce the risk of misjudgment but also more accurately determine the optimal obstacle avoidance timing. The comprehensive consideration of both distance and time dimensions avoids obstacle avoidance that is too early or too late. Moreover, it balances the safety and effectiveness of obstacle avoidance, selecting the best possible time to avoid obstacles while ensuring safety.

[0146] In this application, the step of determining the target longitudinal distance x end Vertical sampling is performed according to preset rule one, specifically as follows:

[0147] With the target longitudinal distance x end Based on the first baseline, longitudinal sampling is performed according to the preset back-off and preset forward sampling intervals.

[0148] The target lateral distance y end Lateral sampling is performed according to preset rule two, specifically as follows:

[0149] With the target lateral distance y end Based on the second reference, lateral sampling is performed according to a preset lateral sampling interval; wherein, when the lateral sampling point exceeds the lane boundary range, the boundary value is taken as the lateral sampling point.

[0150] It should be noted that due to the greater complexity of actual road conditions, a sampling method is still used for route planning. Considering that y has already been calculated through the above steps... end and x end Therefore, this application does not need to use extensive sampling for path planning, but instead uses precise sampling.

[0151] Preferably, the longitudinal sampling method is as follows:

[0152] Based on the obtained x end Based on benchmark one, the sampling compensation is first determined according to the calibration. Typically, 2.5m is taken as the backward sampling step size Interval_back, and 5m is taken as the forward sampling step size Interval_front. In order to maximize the success rate of the sampling path, in addition to adopting x end In addition to longitudinal sampling points, x will be used as the sampling point. end -Interval_back is used as the second sampling point, and x is used as the second sampling point. end +Interval_front, x end +2 * Interval_front are used as the third and fourth sampling points. The above sampling method ensures that, at typical values ​​x... end Sampling was conducted nearby to avoid invalid sampling.

[0153] Preferably, the lateral sampling method is as follows:

[0154] Based on the obtained y end As a baseline, take 1m as the sampling step size Interfal_latt, except for the adopted y end Outside of the horizontal sampling points, with y end -Interval_latt,y end +Interval_latt is used as the second and third sampling points. If the second or third sampling point would cause the vehicle to exceed the lane boundary line of the obstacle avoidance lane, then the maximum and minimum sampling points calculated from the lane boundary line are used as the second and third sampling points.

[0155] The above method reduces the number of sampling points for each path planning iteration to 9-12, minimizing invalid sampling and improving computational efficiency. It also effectively avoids invalid sampling areas, concentrating sampling points in key obstacle avoidance areas and increasing the probability of generating feasible obstacle avoidance paths. Furthermore, symmetrical sampling both forward and backward, and left and right, ensures the rationality of the sampling range. Lane boundary constraints are also considered to avoid generating illegal paths, further guaranteeing the safety and optimization of the obstacle avoidance path.

[0156] It should be noted that this application can also perform safety testing and evaluation on the obstacle avoidance trajectory. This is because the path x calculated in step 1... end Since the lateral acceleration constraint conditions have already been ensured, theoretically there is no need to check the lateral acceleration constraint of the sampling path again. However, considering safety and road uncertainties, such as roads with discontinuous curvature, the curvature and acceleration constraint can be checked again on the sampling path. In this invention, instead of checking every discrete point, only checking the longitudinal distance within x... end The curvature and lateral acceleration at five points: 0%, 20%, 40%, 60%, 80%, and 100%. As mentioned above, the maximum curvature usually occurs at the 20% and 80% positions. Therefore, the above checks can effectively reduce the amount of computation while accurately verifying the maximum curvature point, and are not limited to this.

[0157] In summary, Embodiment 1 of this application accurately calculates the obstacle avoidance start point and path endpoint under extreme obstacle avoidance conditions by comprehensively considering lateral and longitudinal constraints as well as obstacle constraints. It also covers various working scenarios, significantly improving the obstacle avoidance success rate. This application further calculates the target collision time for the minimum obstacle avoidance turn or the target collision distance for the latest obstacle avoidance turn, which is used by the system to determine whether to execute the obstacle avoidance turn function, vehicle instability, and collision risk. Furthermore, based on the accurately calculated lateral and longitudinal target distances for obstacle avoidance turn, this application performs lateral and longitudinal sampling, effectively improving the sampling success rate, reducing the number of samples, and increasing execution efficiency.

[0158] Example 2:

[0159] Please refer to Figure 4 In this second embodiment, a vehicle obstacle avoidance system is also provided based on the above-described vehicle obstacle avoidance method. The system mainly includes a first processing unit, a second processing unit, a third processing unit, and a fourth processing unit.

[0160] The first processing unit is used to respond to obstacle avoidance warning commands and, based on the current preset center position of the vehicle and the longitudinal distance x of the target obstacle to be avoided, performs the following operations. end This allows us to calculate the location of the collision point on one side of the vehicle.

[0161] It should be noted that the first processing unit determines the target's longitudinal obstacle avoidance distance x based on the vehicle's current state and the target's longitudinal obstacle avoidance distance. end This method calculates the location of the collision point on one side of the vehicle. The underlying principle is to establish a vehicle kinematic model and, based on parameters such as current speed and direction, solve equations to determine the coordinates of the collision point on one side of the vehicle under assumed collision conditions.

[0162] The second processing unit is used to obtain the target lateral distance y for obstacle avoidance of the vehicle based on the collision point location on one side of the vehicle and a preset lateral offset ratio of the vehicle. end To obtain the starting point for obstacle avoidance.

[0163] It should be noted that the second processing unit calculates the optimal target lateral distance y by iteratively calculating the collision point location on one side of the vehicle and a preset range of vehicle lateral offset ratios. end The implementation principle involves iterating through multiple sets of parameters to find the y-axis that minimizes the obstacle avoidance trajectory. end .

[0164] The third processing unit is used to calculate the target obstacle avoidance parameters based on the vehicle's current speed, or the vehicle's current speed and the obstacle's current speed, and to output an obstacle avoidance steering command only when the vehicle's current actual obstacle avoidance parameters meet the target obstacle avoidance parameters.

[0165] It should be noted that the third processing unit calculates the time or distance threshold parameters for obstacle avoidance, compares them with the current actual state to determine whether the obstacle avoidance conditions are met, and outputs an obstacle avoidance command. The implementation principle is to establish a kinematic model and calculate the optimal time or distance parameters required for obstacle avoidance.

[0166] The fourth processing unit is configured to respond to the obstacle avoidance steering command and, based on the target longitudinal distance x, end Vertical sampling is performed according to preset rule one, and based on the target lateral distance y end Lateral sampling is performed according to preset rule two to generate obstacle avoidance path planning.

[0167] It should be noted that the fourth processing unit is based on x end and y end Sampling generates obstacle avoidance paths. The principle behind this is to perform longitudinal and lateral sampling at key locations according to preset rules, and then perform curve fitting to obtain the final trajectory.

[0168] Preferably, on a straight road, assume the autonomous vehicle is traveling at 60 km / h. Suddenly, an out-of-control vehicle appears 50 meters ahead and crosses the road.

[0169] At this point, the first processing unit calculates the required longitudinal distance x to the target, which is 40 meters, based on parameters such as the current speed of 60 km / h and the maximum steering angular velocity. endObstacle avoidance is performed, and the coordinates of the collision point (30,0) on the side of the vehicle assuming a collision with the obstacle are calculated. The second processing unit iterates through different lateral offset ratios of the vehicles and calculates y. end The shortest trajectory is found when the distance is 2 meters, thus determining the lateral distance to the target. The third processing unit calculates that an obstacle avoidance distance of 80 meters is required, but the current distance to the obstacle is only 50 meters, which does not meet the obstacle avoidance conditions, so no command is output. At this point, the fourth processing unit does not need to generate an obstacle avoidance path, the system performs emergency braking, and re-evaluates the obstacle avoidance plan after the distance to the vehicle is greater than 80 meters.

[0170] In summary, the vehicle obstacle avoidance system provided in Embodiment 2 of this application achieves accurate calculation of the entire vehicle obstacle avoidance parameters through the first, second, third and fourth processing units, so as to obtain accurate obstacle avoidance path sampling and the optimal start and end points of the obstacle avoidance planning path with higher safety and reliability. The specific implementation process and principle have been described in detail in Embodiment 1, and will not be repeated in this embodiment.

[0171] Example 3:

[0172] This application also provides a vehicle that includes at least the vehicle obstacle avoidance system described above.

[0173] It should be noted that the obstacle avoidance system installed in the vehicle comprises four processing units. The vehicle is equipped with a sensor system capable of detecting road conditions and obstacles ahead. The information collected by the sensors is transmitted to the first processing unit of the obstacle avoidance system. The four processing units work collaboratively according to a pre-defined process to complete obstacle avoidance judgment and provide an obstacle avoidance path. The obstacle avoidance path signal is sent to the vehicle's underlying control system to complete obstacle avoidance actions such as braking and steering.

[0174] Preferably, the vehicles in the embodiments of this application, such as: autonomous vehicles, can be applied to obstacle avoidance decision-making and path planning of autonomous vehicles to improve their obstacle avoidance safety and efficiency; driver assistance systems can be integrated into active safety systems such as automatic emergency braking and automatic steering systems to assist drivers in obstacle avoidance; intelligent transportation systems can be applied to transportation systems such as buses and trucks to assist or replace drivers in obstacle avoidance operations; intelligent logistics vehicles can be applied to unmanned forklifts, handling vehicles, etc. in warehousing and logistics to realize their automatic obstacle avoidance capabilities; intelligent robots can be applied to various intelligent robots that require autonomous obstacle avoidance capabilities, such as service robots and exploration robots; and intelligent toys can be applied to intelligent toy cars and robots that require obstacle avoidance functions, and are not limited thereto.

[0175] Example 4:

[0176] This application also provides a storage medium, which is a type of computer-readable storage medium, having stored thereon a computer program that, when executed by a processor, implements the vehicle obstacle avoidance method as described above.

[0177] The computer-readable storage medium can be configured on any computer device, such as a smartphone, tablet, desktop computer, or cloud server, or it can be an in-vehicle terminal or control system. For example, it may also include input / output devices, network access devices, etc.

[0178] The processor can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The memory is used to store the operating system, application programs, bootloader, data, and other programs, such as the program code of the computer program. The memory can also be used to temporarily store data that has been output or will be output.

[0179] In the several embodiments provided in this application, it will be understood that each block in the flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the figures. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved.

[0180] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0181] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application for those skilled in the art.

Claims

1. A method for vehicle obstacle avoidance, wherein the direction in which the vehicle encounters an obstacle during travel is defined as the longitudinal direction X, and the direction perpendicular to the road surface is defined as the lateral direction Y, characterized in that... include: S1: In response to the obstacle avoidance warning command, based on the current preset center position of the vehicle and the longitudinal distance x of the target obstacle to be avoided by the vehicle. end The location of the collision point on one side of the vehicle is calculated; specifically: When the obstacle avoidance warning command is received, the longitudinal lane change completion rate when the vehicle collides with the obstacle is obtained according to the preset threshold vehicle lateral offset ratio. The process of obtaining the longitudinal lane change completion rate specifically involves: Based on vehicle geometry and a fifth-order polynomial for obstacle avoidance trajectory planning, the relationship between the longitudinal lane change completion degree U and La is constructed as follows: By iterating through La and taking its values, multiple sets of longitudinal lane change completion degrees U are obtained. Where, U=x / x end La represents any point (x) on the side of the vehicle in a potential collision. a ,y a The longitudinal distance La from the preset center position (x, y) of the vehicle is defined as the value of [-Lr, Lf]. Then, based on the target longitudinal distance x end Based on the longitudinal lane change completion degree U, the collision point position on one side of the vehicle is calculated, and the coordinates of the collision point position are (x... c ,y c ); Lr is the distance from the vehicle's preset center position (x, y) to the rear end face of the vehicle, where "-" indicates direction; Lf is the distance from the preset center position (x,y) of the vehicle to the front face of the vehicle; The target longitudinal distance x end It refers to the shortest longitudinal distance, that is, the shortest longitudinal path for a vehicle to change lanes and avoid obstacles under the condition that the surrounding environmental parameters are all permissible; The target lateral distance yend refers to the shortest lateral offset distance from the vehicle's centerline to the obstacle avoidance endpoint. The vehicle's lateral offset ratio is the distance the vehicle's position coordinates at the time of collision with the obstacle moved laterally relative to the obstacle avoidance starting point, divided by the target lateral distance y. end The ratio of the vehicle's lateral offset ratio y ratio =(2y c +B) / 2y end B is the width of the vehicle; S2: Based on the collision point location on one side of the vehicle and a preset threshold vehicle lateral offset ratio, obtain the target lateral distance y for obstacle avoidance by the vehicle. end To obtain the target obstacle avoidance starting point; specifically: The target lateral distance y for vehicle obstacle avoidance end Includes multiple sets of target lateral distances y; Based on the location of the collision point on one side of the vehicle and the lateral offset ratio of the vehicle that varies at preset intervals within a preset threshold range, multiple sets of target lateral distances y for the vehicle obstacle avoidance are calculated. Based on the vehicle speed values ​​at preset intervals and the target lateral distance y of the vehicle obstacle avoidance, multiple sets of target obstacle avoidance starting points with the vehicle lateral offset ratio varying at preset intervals are obtained. Among these, the minimum value among the multiple sets of target lateral distances y is the final target lateral distance y for vehicle obstacle avoidance. end Thus, the longitudinal distance x of the target is determined. end To obtain the target obstacle avoidance starting point; S3: Calculate the target obstacle avoidance parameters based on the vehicle's current speed, or the vehicle's current speed and the obstacle's current speed, and output the obstacle avoidance steering command only when the vehicle's current actual obstacle avoidance parameters meet the target obstacle avoidance parameters; S4: In response to the obstacle avoidance steering command, based on the target longitudinal distance x end Vertical sampling is performed according to preset rule one, and based on the target lateral distance y end Lateral sampling is performed according to preset rule two to generate obstacle avoidance path planning; Prior to step S1, the method further includes: Obtain the curvature value of the vehicle traveling on the current path and the vehicle's current lateral controllable threshold acceleration; Based on the fifth-order polynomial Obstacle avoidance trajectory planning is performed based on the curvature value and the lateral controllable threshold acceleration a. ymax The target longitudinal distance x for obstacle avoidance by the vehicle is calculated. end , .

2. The vehicle obstacle avoidance method according to claim 1, characterized in that, The target obstacle avoidance parameters include the target obstacle avoidance distance; The target obstacle avoidance distance is calculated based on the vehicle's current speed, the obstacle's current speed, and the preset collision compensation distance.

3. The vehicle obstacle avoidance method according to claim 2, characterized in that, The target obstacle avoidance parameters also include the target obstacle avoidance time; The target obstacle avoidance time is calculated based on the vehicle's current speed and the preset collision compensation time. The obstacle avoidance steering command is output only when the vehicle's current actual obstacle avoidance distance is greater than the target obstacle avoidance distance and the vehicle's current actual obstacle avoidance time is greater than the target obstacle avoidance time.

4. The vehicle obstacle avoidance method according to claim 3, characterized in that, The target longitudinal distance x end Vertical sampling is performed according to preset rule one, specifically as follows: With the target longitudinal distance x end Based on the first baseline, longitudinal sampling is performed according to the preset back-off and preset forward sampling intervals; The target lateral distance y end Lateral sampling is performed according to preset rule two, specifically as follows: With the target lateral distance y end Based on the second reference, lateral sampling is performed according to a preset lateral sampling interval; wherein, when the lateral sampling point exceeds the lane boundary range, the boundary value is taken as the lateral sampling point.

5. A vehicle obstacle avoidance system, characterized in that, include: The first processing unit, the second processing unit, the third processing unit, and the fourth processing unit; The first processing unit is used to respond to obstacle avoidance warning commands and, based on the current preset center position of the vehicle and the longitudinal distance x of the target obstacle to be avoided, calculates the obstacle avoidance parameters. end To calculate the location of the collision point on one side of the vehicle; The calculated location of the collision point on one side of the vehicle is specifically as follows: When the obstacle avoidance warning command is received, the longitudinal lane change completion rate when the vehicle collides with the obstacle is obtained according to the preset threshold vehicle lateral offset ratio. The process of obtaining the longitudinal lane change completion rate specifically involves: Based on vehicle geometry and a fifth-order polynomial for obstacle avoidance trajectory planning, the relationship between the longitudinal lane change completion degree U and La is constructed as follows: By iterating through La and taking its values, multiple sets of longitudinal lane change completion degrees U are obtained. Where, U=x / x end La represents any point (x) on the side of the vehicle in a potential collision. a ,y a The longitudinal distance La from the preset center position (x, y) of the vehicle is defined as the value of [-Lr, Lf]. Then, based on the target longitudinal distance x end Based on the longitudinal lane change completion degree U, the collision point position on one side of the vehicle is calculated, and the coordinates of the collision point position are (x... c ,y c ); Lr is the distance from the vehicle's preset center position (x, y) to the rear end face of the vehicle, where "-" indicates direction; Lf is the distance from the preset center position (x,y) of the vehicle to the front face of the vehicle; The target longitudinal distance x end It refers to the shortest longitudinal distance, that is, the shortest longitudinal path for a vehicle to change lanes and avoid obstacles under the condition that the surrounding environmental parameters are all permissible; The target lateral distance yend refers to the shortest lateral offset distance from the vehicle's centerline to the obstacle avoidance endpoint; The vehicle's lateral offset ratio is the distance the vehicle's position coordinates at the time of collision with the obstacle moved laterally relative to the obstacle avoidance starting point, divided by the target lateral distance y. end The ratio of the vehicle's lateral offset ratio y ratio =(2y c +B) / 2y end B is the width of the vehicle; Also includes: Obtain the curvature value of the vehicle traveling on the current path and the vehicle's current lateral controllable threshold acceleration; Based on the fifth-order polynomial Obstacle avoidance trajectory planning is performed based on the curvature value and the lateral controllable threshold acceleration a. ymax The target longitudinal distance x for obstacle avoidance by the vehicle is calculated. end , ; The second processing unit is used to obtain the target lateral distance y for obstacle avoidance of the vehicle based on the collision point location on one side of the vehicle and a preset lateral offset ratio of the vehicle. end To obtain the starting point for obstacle avoidance; The starting point for obtaining the target obstacle avoidance is specifically: The target lateral distance y for vehicle obstacle avoidance end Includes multiple sets of target lateral distances y; Based on the location of the collision point on one side of the vehicle and the lateral offset ratio of the vehicle that varies at preset intervals within a preset threshold range, multiple sets of target lateral distances y for the vehicle obstacle avoidance are calculated. Based on the vehicle speed values ​​at preset intervals and the target lateral distance y of the vehicle obstacle avoidance, multiple sets of target obstacle avoidance starting points with the vehicle lateral offset ratio varying at preset intervals are obtained. Among these, the minimum value among the multiple sets of target lateral distances y is the final target lateral distance y for vehicle obstacle avoidance. end Thus, the longitudinal distance x of the target is determined. end To obtain the target obstacle avoidance starting point; The third processing unit is used to calculate the target obstacle avoidance parameters based on the current vehicle speed, or the current vehicle speed and the current speed of the obstacle, and outputs the obstacle avoidance steering command only when the current actual obstacle avoidance parameters of the vehicle meet the target obstacle avoidance parameters. The fourth processing unit is configured to respond to the obstacle avoidance steering command and, based on the target longitudinal distance x, end Vertical sampling is performed according to preset rule one, and based on the target lateral distance y end Lateral sampling is performed according to preset rule two to generate obstacle avoidance path planning.

6. A vehicle, characterized in that, The vehicle includes at least the vehicle obstacle avoidance system as described in claim 5.

7. A storage medium, one of the computer-readable storage media, characterized in that, It stores a computer program, which, when executed by a processor, implements the vehicle obstacle avoidance method as described in any one of claims 1-4.

Images