Multi-stage parallel parking path planning method for heavy-duty all-wheel steering vehicle

Abstract

This invention relates to a multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles. Based on vehicle parameters and parallel parking space dimensions and road information detected by sensors, a parallel parking scenario for the all-wheel steering vehicle is established. The starting parking point information for vehicle positioning is input, and a designed algorithm is executed to detect existing safe parking paths in real time. Specifically, a series of parking paths are generated using a parameter sampling method and discretized. Road constraints and vehicle dynamics constraints that need to be satisfied during parking are defined. A path evaluation function is designed to evaluate and optimize paths that meet the constraints, thus completing the optimized design of parallel parking paths for multi-segment, multi-steering mode vehicles. This achieves more flexible parking path design for heavy-duty all-wheel steering vehicles, reduces parking space occupied during parking, and improves parking efficiency.

Description

Technical Field

[0001] This invention belongs to the field of heavy vehicle path planning technology, specifically relating to a multi-segment parallel parking path planning method for heavy all-wheel steering vehicles. Background Technology

[0002] Heavy-duty all-wheel steering vehicles are indispensable core engineering equipment in my country's national economic construction and military heavy industry sectors. They are widely used in logistics, engineering construction, and military operations, playing a vital role in improving production efficiency and transportation capacity, and ensuring national security. In recent years, with the surge in the number of heavy-duty multi-axle vehicles, the supply and demand of parking spaces for heavy-duty vehicles in cities has become severely insufficient. To address this, parallel parking methods are being adopted, designating spaces on both sides of safe sections of urban roads as parking areas for heavy-duty vehicles, which alleviates this problem to some extent.

[0003] However, the long body, heavy weight, and high center of gravity of heavy multi-axle vehicles, along with inherent vehicle dynamic constraints, severely limit their maneuverability. This makes parallel parking extremely difficult for them in urban areas. Furthermore, the complexity of parking operations for multi-axle vehicles can easily lead to traffic congestion and even serious traffic accidents due to prolonged parking maneuvers. Therefore, there is a need to propose an automated parallel parking path planning method for heavy all-wheel steering vehicles to improve parking efficiency while ensuring safety.

[0004] Current parallel parking path planning methods employ geometric curve interpolation, using methods such as straight lines and arcs, transition curves, B-spline curves, Bézier curves, and polynomial curves to design paths, or a combination of these methods, to construct multi-segment parking paths to find a collision-free path for parking. For example, patent application 202110248819.X describes using arcs and multi-segment curves to plan continuous curves, considering vehicle motion constraints and collision constraints during parking, and then establishing a parking model. Through several path planning iterations, an objective function is established, and an optimal path is selected as the actual reference parking path for the vehicle, as described in patent application 202210800689.0. These parallel parking path methods help improve the parking efficiency of autonomous vehicles, but they still have some shortcomings and limitations, mainly manifested in:

[0005] 1. Existing parallel parking path planning methods primarily focus on autonomous vehicles with front-axle steering, neglecting the impact of the various steering modes of all-wheel-steering vehicles on current path planning. Furthermore, given the generally longer body and heavier load characteristics of all-wheel-steering vehicles, using existing parallel parking path planning methods will result in excessive parking space occupation and low parking efficiency. Therefore, it is necessary to study the differences in dynamic performance of different steering modes for all-wheel-steering vehicles and adopt parking path planning methods with more flexible behavior based on different steering modes to improve parking efficiency.

[0006] 2. Existing parallel parking path planning methods require high precision in locating the parking start point. When there is an error between the actual parking point and the reference parking point, parking failure is likely. When the parking point cannot be accurately located, the vehicle needs to repeatedly move to approach it, further exacerbating the low efficiency of parallel parking for heavy multi-axle vehicles. Current solutions utilize vehicle sensors, cameras, and LiDAR for real-time measurement, along with more precise motion control algorithms, to achieve accurate positioning at the preset parking point. However, this technology is difficult to implement and increases costs, making it difficult to apply to heavy-duty all-wheel-steering vehicles. Therefore, this paper proposes an improvement to the traditional path planning method by incorporating a dynamic starting parking point design to enhance the robustness of the parking start point. Summary of the Invention

[0007] The purpose of this invention is to provide a multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles, so as to improve the safety and efficiency of parallel parking of heavy-duty all-wheel steering vehicles.

[0008] To achieve the above objectives, the technical solution of the present invention is: a multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles, comprising the following steps:

[0009] Step S1: Design the vehicle shape module based on the parameters of the heavy-duty all-wheel steering vehicle, as well as the parallel parking space size and road information detected by the sensors, to establish a parallel parking scenario for the heavy-duty all-wheel steering vehicle.

[0010] Step S2: Using the method of directly giving the starting parking point, drive the vehicle to the parking position to complete the pre-parking preparation. Then, execute the designed multi-segment path planning algorithm to detect in real time whether there is a safe parallel parking path with the current vehicle pose as the starting parking point.

[0011] Step S3: Treat parking as the reverse process of a vehicle leaving the parking space, and design the parking path in three segments. The first segment completes the task of reversing the vehicle to a safe distance behind the parking space. The second segment adopts a crab-like mode, with all wheels turning at the same angle to achieve collision-free lateral parking. The third segment adopts a minimum radius mode, using a fifth-order polynomial curve to design a continuous curvature path. The three segments work together to complete the exit movement.

[0012] Step S4: Generate the second path segment according to the parameter sampling method, design the third path segment based on the second path segment, combine it with the first path segment to obtain a series of path clusters, extract a series of discrete points in the designed path, calculate the X and Y positions, driving distance and path curvature information of each point, and store the data.

[0013] Step S5: Establish the conditions for road constraints and vehicle dynamic curvature constraints, process the obtained series of path clusters, and filter out the paths that satisfy the constraints.

[0014] Step S6: Design a path evaluation function to evaluate and select the best path that meets the constraints. This path is then used as the optimal parking path. The heading angle of the vehicle and the turning angle of each wheel are calculated for each process and input into the established vehicle shape module to complete the planning of parallel parking motion.

[0015] In one embodiment of the present invention, a road environment and parallel parking space are designed, and a vehicle shape program module is added on the basis to establish a parallel parking scenario for a heavy-duty all-wheel steering seven-axle vehicle. The vehicle shape module includes the method for establishing the vehicle and tire shapes as follows:

[0016] Vehicle body shape creation:

[0017]

[0018] Establishing the shapes of the left and right tires:

[0019]

[0020]

[0021]

[0022] In the formula, A(x,y), B(x,y), C(x,y), and D(x,y) represent four external shape points of the vehicle; tyre_il(x,y) is the steering center of the left tire on the i-th axis; tyre_ilA(x,y), tyre_ilB(x,y), tyre_ilC(x,y), and tyre_ilD(x,y) are the corresponding external shape points of the left tire; tyre_ir(x,y) is the steering center of the right tire on the i-th axis; tyre_irA(x,y), tyre_irB(x,y), tyre_irC(x,y), and tyre_irD(x,y) are the corresponding external shape points of the right tire; L is the vehicle length; W is the vehicle width; X and Y are the coordinates of the vehicle's center of mass; L f L r These represent the front and rear overhang distances of the vehicle, respectively; l is the tire length; and w is the tire width. i Let θ be the vector distance between the centroid and the i-th axis, θ be the vehicle yaw angle, and δ be the vector distance between the centroid and the i-th axis. i Let be the rotation angle of the tire on the i-th axis.

[0023] In one embodiment of the present invention, at the vehicle pre-parking point, the coordinate information obtained by actual measurement is used as the input of a multi-segment path planning algorithm. Parking is equivalent to the reverse process of vehicle exiting the parking space and is divided into three parking paths for design. The algorithm detects whether there is a safe parking path. If it shows that there is no safe parking path, the vehicle position is adjusted until a safe parallel parking path is planned.

[0024] In one embodiment of the present invention, the first path completes the task of reversing the vehicle to a predetermined safe distance behind the parking space, leaving more room for subsequent path design. Its path planning equation is:

[0025]

[0026] In the formula, P0(X0,Y0) is the expected endpoint after safe parking, P1(X1,Y1) is the endpoint coordinate of the first path segment, and L is the vehicle length. safe To maintain a safe distance from the back of the parking space, L p This is the length of the parking space.

[0027] In one embodiment of the present invention, the second path segment adopts a crab-like mode. The parameter sampling method is used to sample the two parameters, the turning angle and the distance, in this mode to obtain different second path segments. The path planning equation is as follows:

[0028]

[0029] K2∈(K 2min ,K 2max )

[0030]

[0031] K 2max =tan(δ) max )

[0032] In the formula, P1(X1,Y1) is the starting coordinate of the second path segment, P2(X2,Y2) is the ending coordinate of the second path segment, K2 and S2 are the slope and travel distance of the crab-shaped path, respectively, and K... 2min K 2max L represents the minimum and maximum slopes of the crab-pattern path, respectively. p W is the length of the parking space. p δ is the width of the parking space. max This is the maximum steering angle of the wheel.

[0033] In one embodiment of the present invention, the third path segment adopts a minimum radius mode. Based on the given starting parking point information P3(X3,Y3) as the endpoint information of the third path segment and the endpoint information P2(X2,Y2) of the second path segment as the starting point information of the third path segment, the path equation with continuous curvature is designed using a fifth-order polynomial curve:

[0034]

[0035] In the formula, P2(X2,Y2) is the starting coordinate of the third path segment, P3(X3,Y3) is the ending coordinate of the third path segment, a0~a5 are the parameters of the fifth-order polynomial curve, θ2 and θ3 are the yaw angles of the vehicle at the starting and ending positions of the third path segment, and L1 is the distance between the vehicle's first axle and the center of mass. This is equivalent to the first axle rotation angle of the vehicle at the starting and ending coordinate points of the third path.

[0036] In one embodiment of the present invention, a series of discrete points for the three designed paths are extracted, and the X and Y positions, travel distances, and path curvature information of the discrete points are calculated and stored. Road constraints and vehicle dynamics curvature constraints are established, and the resulting series of path clusters are processed to select safe paths that meet the constraints. The following is a detailed process of path discretization:

[0037]

[0038]

[0039]

[0040] In the formula N 01 N 12 N 23 To set the number of discrete points for each path segment, [X]01 ,Y 01 ]、[X 12 ,Y 12 ]、[X 23 ,Y 23 [S represents the X and Y position information of each discrete point] 01 S 12 S 23 Let θ be the distance traveled on each segment of the path. 01 θ 12 θ 23 Let ρ be the yaw angle of each path segment. 01 ρ 12 ρ 23 Let P0(X0,Y0) be the path curvature of each segment, P1(X1,Y1) be the expected endpoint after safe parking, P2(X2,Y2) be the endpoint coordinates of the first segment, P3(X3,Y3) be the endpoint coordinates of the second segment, and P3(X3,Y3) be the endpoint coordinates of the third segment. K2 and S2 are the slope and driving distance of the crab-shaped path, and a0 to a5 are the parameters of the fifth-order polynomial curve.

[0041] In one embodiment of the present invention, the vehicle information and path information after each path are discretized are stored in a cell array, and the implementation method is as follows:

[0042] X{m}=[X 01 ,X 12 ,X 23 ]

[0043] Y{m}=[Y 01 ,Y 12 ,Y 23 ]

[0044] S{m}=[S 01 ,S 12 ,S 23 ]

[0045] θ{m}=[θ 01 ,θ 12 ,θ 23 ]

[0046] ρ{m}=[ρ 01 ,ρ 12 ,ρ 23 ]

[0047]

[0048] In the formula, X{m}, Y{m}, S{m}, θ{m}, and ρ{m} represent the X and Y coordinates of the path, the distance traveled, the yaw angle, and the path curvature, respectively; A{m}(x,y), B{m}(x,y), C{m}(x,y), and D{m}(x,y) represent the vehicle's external shape; m is the total number of paths; L is the vehicle's length; W is the vehicle's width; X and Y are the coordinates of the vehicle's center of mass; and L... f L r These are the front overhang distance and rear overhang distance of the vehicle, respectively, and θ is the vehicle yaw angle.

[0049] In one embodiment of the present invention, the information in each cell array is analyzed to determine whether the path satisfies road constraints and curvature constraints. The process is as follows:

[0050] Road constraints are:

[0051] road_boundary1 = -Wp / 2

[0052] road_boundary2 = -Wp / 2 + d

[0053]

[0054] The curvature constraint is:

[0055]

[0056]

[0057] In the formula, road_boundary1 is the lower boundary road constraint, road_boundary2 is the upper boundary road constraint, and W p Let be the width of the parking space, d be the total width of the road, A{m}(y), B{m}(y), C{m}(y), D{m}(y) be the Y-coordinate information of the vehicle's external coordinate points, ρ be the path curvature, and R be the vehicle's turning radius. min This is the vehicle's minimum turning radius. This information is obtained for a seven-axle vehicle in all-wheel steering mode at a speed of 2 m / s and turning at the maximum angle.

[0058] In one embodiment of the present invention, the slope, driving distance, total driving distance required by the vehicle during parking, maximum path curvature, and maximum lateral position occupied in the second segment of the crab-walking mode are selected as evaluation factors. An evaluation function is designed to evaluate the path that meets the constraints. The process of establishing the evaluation function is as follows:

[0059]

[0060] Where n represents the total number of paths that satisfy the constraints, K2{n} and S2{n} are two evaluation factors, namely the slope and driving distance of the crab-patterned parking path after the constraints have been filtered, respectively. max {n}、ρ max {n}、Y max {n} represents the total driving distance for parking, the maximum path curvature, and the maximum lateral position occupied, respectively. Q1, Q2, Q3, Q4, and Q5 represent their respective weights.

[0061] Compared with the prior art, the present invention has the following beneficial effects:

[0062] 1) This invention solves the problem of large space occupation and low efficiency when heavy-duty all-wheel steering vehicles perform parallel parking operations. Addressing the shortcomings of existing parallel parking path planning methods, an improved multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles is proposed. Considering the dynamic performance advantages of the multiple steering modes possessed by all-wheel steering vehicles, a more flexible parking path planning approach is designed, reducing the parking space occupied by heavy-duty all-wheel steering vehicles during parallel parking while improving parking efficiency.

[0063] 2) This invention addresses the problem of reduced parking efficiency or even failure caused by the high precision requirement of the reference parking starting point in parallel parking path planning methods. When the vehicle cannot be accurately positioned at the designated parking point, repeated maneuvers to approach the parking point are necessary, leading to low parking efficiency. Large positioning errors at the parking starting point can also cause collisions during parking. The method of this invention uses real-time vehicle information as a parking reference point during the parking preparation stage to determine if a safe parking path exists, eliminating the need to return to the parking starting point obtained by general parallel parking path planning algorithms. This improves the robustness of the parking path starting point and also increases the efficiency of parallel parking for heavy-duty all-wheel-steering vehicles. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of a parallel parking scenario for a heavy-duty all-wheel steering vehicle in an embodiment of the present invention.

[0065] Figure 2 This is a schematic diagram of the multi-segment parallel parking path design process for all-wheel steering vehicles in an embodiment of the present invention.

[0066] Figure 3 This is a schematic diagram of the multi-segment parallel parking path design for an all-wheel steering vehicle in an embodiment of the present invention.

[0067] Figure 4 This is a schematic diagram showing the relationship between the turning radius and the turning angle in the minimum radius mode of this invention embodiment.

[0068] Figure 5 This is a schematic diagram of sampling and planning parking paths at different parking starting points in an embodiment of the present invention.

[0069] Figure 6 This is a schematic diagram of the upper and lower road boundary constraints in an embodiment of the present invention.

[0070] Figure 7 This is a schematic diagram of the paths that meet the constraints selected in the embodiments of the present invention.

[0071] Figure 8 This is a diagram showing the optimal path obtained by the method of this invention at different parking points. Detailed Implementation

[0072] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings.

[0073] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application.

[0074] Figure 1 This is a schematic diagram of a parallel parking scenario for a heavy-duty all-wheel steering multi-axle vehicle in this embodiment. Taking a seven-axle all-terrain crane as an example, the first three axles of this crane use a mechanical linkage steering system, while the last four axles use an electro-hydraulic steering system. Therefore, it can achieve independent steering for each axle and has multiple steering modes such as crab steering, minimum radius steering, front axle steering, and center axle locking without steering. Based on the relevant design of the road and parking space dimensions, the design length is L. p Width is W p The parking space has a road width of d, and a vehicle shape program module is added to create a parallel parking scenario for a heavy-duty all-wheel steering seven-axle vehicle.

[0075] Figure 2 This is a schematic diagram of the multi-segment parallel parking path design process for all-wheel steering vehicles in an embodiment of the present invention. It shows the overall multi-segment path design method and the overall implementation process for determining whether a path that meets safety constraints exists at the parking point.

[0076] Figure 3 This is a schematic diagram of a multi-segment parallel parking path design for an all-wheel steering vehicle in an embodiment of the present invention. The first segment completes the reversing task; the second segment switches to crab mode to park laterally to a suitable position; and the third segment uses the minimum radius mode to drive along a designed fifth-order polynomial curve to the given starting parking point.

[0077] Figure 4 This is a schematic diagram illustrating the relationship between the turning radius and the steering angle in the minimum radius mode of this invention embodiment. In particular, it shows the geometric relationship between the turning radius and the first axis steering angle, and the connection between the yaw angle and the fifth-degree polynomial equation, thereby calculating the parameters a0 to a5 in the fifth-degree polynomial equation.

[0078] Figure 5 This is a schematic diagram of parking path planning based on different parking starting points in an embodiment of the present invention. To address the issue of parking point robustness, the algorithm can generate parallel parking paths based on any given parking starting point. Here, we take six different starting parking points (15,3), (15,4), (15,5), (16,3), (16,4), and (16,5) as examples for analysis to determine whether there exists a safe parking path that satisfies the constraints.

[0079] Figure 6 This is a schematic diagram illustrating the upper and lower road boundary constraints in an embodiment of the present invention. Vehicles should avoid collisions with road boundaries and parking space boundaries during parking.

[0080] Figure 7 This is a schematic diagram of paths that meet the road and vehicle dynamics constraints selected in this embodiment of the invention. By changing the values ​​of parameters K2 and S2 through sampling, a series of path clusters are obtained. Paths that do not meet the road curvature constraints are excluded, and all paths that meet the constraints are selected.

[0081] Figure 8 This diagram illustrates the optimal path obtained by the method of this invention under different starting parking points. An evaluation function is designed to evaluate roads that meet the constraints, and the path with the best evaluation is selected as the optimal path.

[0082] In this embodiment, the following parameters are included for the road parking scenario and the seven-axle vehicle:

[0083]

[0084] This embodiment provides a path planning method to improve the parallel parking efficiency of heavy-duty all-wheel steering vehicles, including the following steps:

[0085] Step S1: Design the vehicle shape module based on the parameters of the heavy-duty all-wheel steering vehicle, as well as the parallel parking space size and road information detected by the sensors, to establish a parallel parking scenario for the heavy-duty all-wheel steering vehicle.

[0086] Step S2: Using the method of directly specifying the starting parking point, when preparing to park, drive the vehicle to a suitable parking position and start real-time detection to see if there is a parallel parking safety path with the current vehicle position as the starting parking point.

[0087] Step S3: Using an exit-from-the-parking approach, the parking path is divided into three segments for design. The first segment involves the vehicle reversing backwards to a predetermined safe distance behind the parking space without turning. The second segment uses a crab-like maneuver, where all wheels turn at the same angle to ensure the vehicle exits the parking space without collision in the lateral direction. The third segment uses a minimum radius design, employing a fifth-order polynomial curve to create a path with continuous curvature.

[0088] Step S4: Generate the second path segment according to the parameter sampling method, design the third path segment based on the second path segment, obtain a series of path clusters, divide the designed path into a series of discrete points, and calculate the X and Y positions, driving distance, path curvature and other information of each point.

[0089] Step S5: Establish road constraints and vehicle dynamics curvature constraints, process a series of path clusters containing path information, and filter out paths that meet the constraints;

[0090] Step S6: Design a path evaluation function to evaluate and select the best path that meets the constraints. This path is then used as the optimal parking path. The heading angle of the vehicle and the turning angle of each wheel are calculated in each process and input into the established vehicle shape model to complete the planning of parallel parking motion.

[0091] In step S1, based on the relevant design of the road and parking space dimensions, the designed length is L. p Width is W p Given a parking space and a road with a width of d, design the exterior module for a seven-axle vehicle and create a parallel parking scenario for the seven-axle vehicle. The vehicle exterior module includes the methods for creating the vehicle's overall shape and tire shapes as follows:

[0092]

[0093] The steering center of the i-th axis tire and the shapes of the left and right tires are established:

[0094]

[0095]

[0096]

[0097] In the formula, A(x,y), B(x,y), C(x,y), and D(x,y) represent four external shape points of the vehicle; tyre_il(x,y) is the steering center of the left tire on the i-th axis; tyre_ilA(x,y), tyre_ilB(x,y), tyre_ilC(x,y), and tyre_ilD(x,y) are the corresponding external shape points of the left tire; tyre_ir(x,y) is the steering center of the right tire on the i-th axis; tyre_irA(x,y), tyre_irB(x,y), tyre_irC(x,y), and tyre_irD(x,y) are the corresponding external shape points of the right tire; L is the vehicle length; W is the vehicle width; X and Y are the coordinates of the vehicle's center of mass; L f L r These represent the front and rear overhang distances of the vehicle, respectively; l is the tire length; and w is the tire width. i Let θ be the vector distance between the centroid and the i-th axis, θ be the vehicle yaw angle, and δ be the vector distance between the centroid and the i-th axis. i Let be the rotation angle of the tire on the i-th axis.

[0098] In step S2, starting from the given actual parking location, the path planning algorithm first checks whether there is a safe parking path. If it shows that there is no safe parking path, the vehicle position is adjusted until a parallel parking path is planned.

[0099] In step S3, during the path design phase, the movement of parking the vehicle is often considered equivalent to the reverse process of exiting the parking space. Therefore, an exit-oriented approach is adopted, and the path is divided into three segments for separate design. The first segment completes the task of the vehicle reversing in a straight line without turning to a safe distance behind the parking space, leaving more operating space. The second segment uses a crab-like mode, where all wheels turn at the same angle to achieve collision-free lateral exit from the parking space, which is crucial for successful parking. Since the turning angle and distance during the crab-like mode have a significant impact on the driving path, a sampling method is used to control these two parameters, resulting in different second-segment paths. The third segment uses a minimum radius mode, using the given starting parking point information P3(X3,Y3) as the endpoint information of the third segment and the endpoint information P2(X2,Y2) of the second segment as the starting point information of the third segment, and a path with continuous curvature is designed using a fifth-order polynomial curve.

[0100]

[0101]

[0102]

[0103] K2∈(K 2min ,K 2max )

[0104]

[0105] K 2max =tan(δ) max )

[0106] In the formula, P0(X0,Y0) is the expected endpoint after safe parking, P1(X1,Y1), P2(X2,Y2), and P3(X3,Y3) are the coordinates of each segment of the path, and L... safe K2 and S2 represent the safe distance from the rear of the parking space, respectively. K2 and S2 are the slope and travel distance of the second crab-shaped path, respectively, and their values ​​can be calculated using the above formula. θ2 and θ3 are the yaw angles of the vehicle at the two positions. Equivalent to the first axle rotation angle of the vehicle at two coordinate points, a0~a5 are the parameters of the fifth-order polynomial curve, δ max This is the maximum steering angle of the wheel.

[0107] In step S4, since the second path segment in step S3 was generated based on parameter sampling and affected the design of the third path segment, a series of path clusters can be obtained. The three designed paths are divided into a series of discrete points, and information such as the X and Y positions, driving distance, and path curvature of the discrete points are calculated. The following is a detailed process of path discretization.

[0108]

[0109]

[0110]

[0111] Where N 01 N 12 N 23 To set the number of discrete points for each path segment, [X] 01 ,Y 01 ]、[X 12 ,Y 12 ]、[X 23 ,Y 23 [S represents the X and Y position information of each discrete point] 01 S 12 S 23 Let θ be the distance traveled on each segment of the path. 01 θ 12 θ 23 Let ρ be the yaw angle of each path segment. 01 ρ 12 ρ 23 Let be the path curvature of each path segment.

[0112] In step S5, the discretized information of each path is stored in cell arrays X{m}, Y{m}, S{m}, θ{m}, and ρ{m}, where m is the total number of paths. The information in each cell array is analyzed. Based on the vehicle shape formula in step 1, the vehicle shape coordinates A(x,y), B(x,y), C(x,y), and D(x,y) corresponding to all discrete points, along with the stored curvature information, are calculated to determine whether the path satisfies road constraints and curvature constraints. Road constraints require that the vehicle body does not collide with road boundaries or parking space boundaries. Curvature constraints require that the curvature of the path be less than the maximum allowable curvature for vehicle movement. When the path clusters planned from the selected starting parking point do not meet the constraints, the console will display that no road meets the conditions.

[0113] The cell array information is as follows:

[0114] X{m}=[X 01 ,X 12 ,X 23 ]

[0115] Y{m}=[Y 01 ,Y 12 ,Y 23 ]

[0116] S{m}=[S 01 ,S 12 ,S 23 ]

[0117] θ{m}=[θ 01 ,θ 12 ,θ 23 ]

[0118] ρ{m}=[ρ 01 ,ρ 12 ,ρ 23 ]

[0119] Road constraints are:

[0120]

[0121] road_boundary1 = -W p / 2

[0122] road_boundary2 = -W p / 2+d

[0123] The curvature constraint is:

[0124]

[0125]

[0126] In the formula, road_boundary1 is the lower boundary road constraint, road_boundary2 is the upper boundary road constraint, and W p Let d be the width of the parking space, d be the total width of the road, A{m}(y), B{m}(y), C{m}(y), D{m}(y) be the Y-coordinate information of the vehicle's external coordinate points, ρ be the path curvature, and R be the vehicle's turning radius. min This is the vehicle's minimum turning radius. This information is obtained for a seven-axle vehicle in all-wheel steering mode at a speed of 2 m / s and turning at its maximum angle.

[0127] In step S6, the slope, driving distance, total driving distance required by the vehicle during the parking process, maximum path curvature, and maximum lateral position occupied in the second segment of the crab mode are selected as evaluation factors. An evaluation function is designed to evaluate the paths that meet the constraints. The weights are adjusted according to the path quality, the function value of each path is calculated, and the path with the best evaluation is selected as the optimal parking path.

[0128]

[0129] Where n represents the total number of paths that satisfy the constraints, K2{n} and S2{n} are two evaluation factors, namely the slope and driving distance of the crab-patterned parking paths after the constraints have been filtered, respectively. max {n}、ρ max {n}、Y max {n} represents the total driving distance for parking, the maximum path curvature, and the maximum lateral position occupied, respectively. Q1, Q2, Q3, Q4, and Q5 represent their respective weights.

[0130] According to this method, considering the performance advantages of heavy-duty all-wheel steering vehicles with different steering modes, a multi-segment curve is designed to achieve a more flexible parking path design, reduce the parking space occupied during the parking process, and improve parking efficiency.

[0131] The above description, including numerical values ​​and definitions, is merely a preferred embodiment of the present invention and is not intended to limit the invention in any other way. Any person skilled in the art may make changes or modifications to the disclosed technical content to create equivalent embodiments. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention, without departing from the scope of the present invention, shall still fall within the protection scope of the present invention.

Claims

1. A multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles, characterized in that, Includes the following steps: Step S1: Design the vehicle shape module based on the parameters of the heavy-duty all-wheel steering vehicle, as well as the parallel parking space size and road information detected by the sensors, to establish a parallel parking scenario for the heavy-duty all-wheel steering vehicle. Step S2: Using the method of directly giving the starting parking point, drive the vehicle to the parking position to complete the pre-parking preparation. Then, execute the designed multi-segment path planning algorithm to detect in real time whether there is a safe parallel parking path with the current vehicle pose as the starting parking point. Step S3: Treat parking as the reverse process of a vehicle leaving the parking space, and design the parking path in three segments. The first segment completes the task of reversing the vehicle to a safe distance behind the parking space. The second segment adopts a crab-like mode, with all wheels turning at the same angle to achieve collision-free lateral parking. The third segment adopts a minimum radius mode, using a fifth-order polynomial curve to design a continuous curvature path. The three segments work together to complete the exit movement. Step S4: Generate the second path segment according to the parameter sampling method, design the third path segment based on the second path segment, combine it with the first path segment to obtain a series of path clusters, extract a series of discrete points in the designed path, calculate the X and Y positions, driving distance and path curvature information of each point, and store the data. Step S5: Establish the conditions for road constraints and vehicle dynamic curvature constraints, process the obtained series of path clusters, and filter out the paths that satisfy the constraints. Step S6: Design a path evaluation function to evaluate and select the best path that meets the constraints. This path is then used as the optimal parking path. The heading angle of the vehicle and the turning angle of each wheel are calculated in each process and input into the established vehicle shape module to complete the planning of parallel parking motion. At the vehicle's pre-parking point, the coordinate information obtained from actual measurements is used as the input of a multi-segment path planning algorithm. Parking is equivalent to the reverse process of vehicle exiting the parking space and is divided into three parking paths for design. The algorithm checks whether there is a safe parking path. If it shows that there is no safe parking path, the vehicle position is adjusted until a safe parallel parking path is planned. The second path segment adopts a crab-like movement pattern. A parameter sampling method is used to sample the two parameters, turning angle and distance, in this pattern to obtain different second path segments. The path planning equation is as follows: In the formula, P1(X1, Y1) is the starting point coordinate of the second path, P2(X2, Y2) is the end point coordinate of the second path, K2 and S2 are the slope and the running distance of the crab mode path, K 2min , K 2max are the minimum and maximum slopes of the crab mode path respectively, L p is the length of the parking space, W p is the car width of the parking space, and δ max is the maximum steering angle of the wheel. The third path segment adopts the minimum radius mode. Based on the given starting parking point information P3(X3,Y3) as the endpoint information of the third path and the endpoint information P2(X2,Y2) of the second path as the starting point information of the third path, the path equation with continuous curvature is designed using a fifth-order polynomial curve: In the formula, P2(X2,Y2) is the starting coordinate of the third path segment, P3(X3,Y3) is the ending coordinate of the third path segment, a0~a5 are the parameters of the fifth-order polynomial curve, θ2 and θ3 are the yaw angles of the vehicle at the starting and ending positions of the third path segment, and L1 is the distance between the vehicle's first axle and the center of mass. This is equivalent to the first axle rotation angle of the vehicle at the starting and ending coordinate points of the third path.

2. The multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles according to claim 1, characterized in that, Design a road environment and parallel parking spaces, and add a vehicle shape programming module to create a parallel parking scenario for a heavy-duty all-wheel steering seven-axle vehicle. The vehicle shape module includes the method for creating the vehicle and tire shapes as follows: Vehicle body shape creation: Establishing the shapes of the left and right tires: In the formula, A(x,y), B(x,y), C(x,y), and D(x,y) represent four external shape points of the vehicle; tyre_il(x,y) is the steering center of the left tire on the i-th axis, tyre_ilA(x,y), tyre_ilB(x,y), tyre_ilC(x,y), and tyre_ilD(x,y) are the corresponding external shape points of the left tire; tyre_ir(x,y) is the steering center of the right tire on the i-th axis, tyre_irA(x,y), tyre_irB(x,y), tyre_irC(x,y), and tyre_irD(x,y) are the corresponding external shape points of the right tire; L is the vehicle length; W is the vehicle width; X and Y are the coordinates of the vehicle's center of mass; L f L r These represent the front and rear overhang distances of the vehicle, respectively; l is the tire length; and w is the tire width. i Let θ be the vector distance between the centroid and the i-th axis, θ be the vehicle yaw angle, and δ be the vector distance between the centroid and the i-th axis. i Let be the rotation angle of the tire on the i-th axis.

3. The multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles according to claim 1, characterized in that, The first segment of the path completes the task of reversing the vehicle to a predetermined safe distance behind the parking space, leaving more room for subsequent path design. Its path planning equation is: In the formula, P0(X0,Y0) is the expected endpoint after safe parking, P1(X1,Y1) is the endpoint coordinate of the first path segment, and L is the vehicle length. safe To maintain a safe distance from the back of the parking space, L p This refers to the length of the parking space.

4. The multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles according to claim 1, characterized in that, A series of discrete points are extracted from the three designed paths, and the X and Y positions, travel distances, and path curvature information of the discrete points are calculated and stored. The conditions for road constraints and vehicle dynamics curvature constraints are established, and the resulting series of path clusters are processed to select safe paths that meet the constraints. The following is a detailed process of path discretization: In the formula N 01 N 12 N 23 To set the number of discrete points for each path segment, [X] 01 ,Y 01 ]、[X 12 ,Y 12 ]、[X 23 ,Y 23 [S represents the X and Y position information of each discrete point] 01 S 12 S 23 Let θ be the distance traveled on each segment of the path. 01 θ 12 θ 23 The yaw angle for each segment of the path. Let P0(X0,Y0) be the path curvature of each segment, P1(X1,Y1) be the expected endpoint after safe parking, P2(X2,Y2) be the endpoint coordinates of the first segment, P3(X3,Y3) be the endpoint coordinates of the second segment, and P3(X3,Y3) be the endpoint coordinates of the third segment. K2 and S2 are the slope and driving distance of the crab-shaped path, and a0~a5 are the parameters of the fifth-order polynomial curve.

5. The multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles according to claim 4, characterized in that, The vehicle information and path information after each path is discretized are stored in a cell array respectively. The implementation method is as follows: In the formula, X{m}, Y{m}, S{m}, θ{m}, These represent the X and Y coordinates of the path, the distance traveled, the yaw angle, and the path curvature. A{m}(x, y), B{m}(x, y), C{m}(x, y), and D{m}(x, y) represent the vehicle's external dimensions, m represents the total number of paths, L represents the vehicle's length, W represents the vehicle's width, X and Y represent the vehicle's center of gravity coordinates, and L represents the vehicle's center of gravity coordinates. f L r These are the front overhang distance and rear overhang distance of the vehicle, respectively, and θ is the vehicle yaw angle.

6. The multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles according to claim 4, characterized in that, The information in each cell array is analyzed to determine whether the path satisfies the road constraint and curvature constraint. The process is as follows: Road constraints are: The curvature constraint is: In the formula, road_boundary1 is the lower boundary road constraint, road_boundary2 is the upper boundary road constraint, and W p Let d be the width of the parking space, d be the total width of the road, and A{m}(y), B{m}(y), C{m}(y), and D{m}(y) be the Y-coordinate information of the vehicle's external coordinate points. R is the path curvature, and R is the turning radius of the vehicle. min This is the vehicle's minimum turning radius. This information is obtained for a seven-axle vehicle in all-wheel steering mode at a speed of 2 m / s and turning at the maximum angle.

7. The multi-segment parallel parking path planning method for heavy-duty all-wheel steering vehicles according to claim 1, characterized in that, The path satisfying the constraints is evaluated, and the slope, travel distance, total travel distance required for parking, maximum path curvature, and maximum lateral position occupied in the second segment of the crab-walking mode are selected as evaluation factors. The evaluation function is established as follows: Where n represents the total number of paths that satisfy the constraints, K2{n} and S2{n} are two evaluation factors, namely the slope and driving distance of the crab-patterned parking path after the constraints have been filtered, respectively. max {n}、 Y max {n} represents the total driving distance for parking, the maximum path curvature, and the maximum lateral position occupied, respectively. Q1, Q2, Q3, Q4, and Q5 represent their respective weights.

Images