Path planning method, electronic device, vehicle, storage medium and program product

Abstract

The invention discloses a path planning method, electronic equipment, a vehicle, a non-volatile computer readable storage medium and a computer program product. The path planning method is applied to the vehicle and comprises the steps that path planning is carried out based on state information of the vehicle and road information to generate a target path, and when the vehicle runs on the target path, the state information of the vehicle meets a preset dynamic cost requirement and / or a chassis trafficability requirement. Thus, path planning can be carried out by comprehensively considering the geographical environment of the current driving environment and the current state of the vehicle, it is ensured that the target path obtained through path planning can meet the dynamic cost requirement and / or the chassis trafficability requirement, and therefore reasonable path planning is achieved; and better trafficability and safety are provided for the unmanned vehicle in a cross-country environment.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and more particularly, to a path planning method, an electronic device, a vehicle, a non-volatile computer-readable storage medium, and a computer program product. Background Art

[0002] Off-road path planning involves designing a safe, efficient, and adaptable path for unmanned vehicles or robots in complex and changing terrain and road conditions. Off-road terrain is complex and unpredictable. Improper path planning can hinder vehicle operation and even compromise safety. For example, choosing the wrong path can create a risk of rollover when navigating a slope. Therefore, optimal path planning is a pressing issue. Summary of the Invention

[0003] Embodiments of the present application provide a path planning method, an electronic device, a vehicle, a non-volatile computer-readable storage medium, and a computer program product.

[0004] The path planning method of the embodiment of the present application is applied to a vehicle, and the method includes: performing path planning based on the vehicle's state information and road information to generate a target path. When the vehicle is traveling on the target path, the vehicle's state information meets preset dynamic cost requirements and / or chassis passability requirements.

[0005] An electronic device according to an embodiment of the present application includes a processor, a memory, and a computer program, wherein the computer program is stored in the memory and executed by the processor, and the computer program includes instructions for executing a path planning method. The path planning method is applied to a vehicle and includes: performing path planning based on the vehicle's state information and road information to generate a target path, wherein when the vehicle travels along the target path, the vehicle's state information satisfies a preset dynamic cost requirement and / or chassis passability requirement.

[0006] A vehicle according to an embodiment of the present application includes an electronic device, the electronic device including a processor, a memory, and a computer program, wherein the computer program is stored in the memory and executed by the processor, and the computer program includes instructions for executing a path planning method. The path planning method is applied to a vehicle, and the method includes: performing path planning based on the vehicle's state information and road information to generate a target path, wherein when the vehicle travels along the target path, the vehicle's state information satisfies a preset dynamic cost requirement and / or chassis passability requirement.

[0007] A non-volatile computer-readable storage medium according to an embodiment of the present application includes a computer program that, when executed by a processor, causes the processor to perform a path planning method. The path planning method is applied to a vehicle and includes: performing path planning based on vehicle state information and road information to generate a target path, wherein when the vehicle travels along the target path, the vehicle state information satisfies a preset dynamic cost requirement and / or chassis passability requirement.

[0008] A computer program product according to an embodiment of the present application includes a computer program configured to execute instructions for a path planning method. The path planning method is applied to a vehicle and includes performing path planning based on vehicle state information and road information to generate a target path, wherein when the vehicle travels along the target path, the vehicle state information satisfies a preset dynamic cost requirement and / or chassis passability requirement.

[0009] The path planning method, electronic device, vehicle, non-volatile computer-readable storage medium, and computer program product of the embodiments of the present application can, based on road information and vehicle status information, determine a target path that ensures that the vehicle's status information meets preset dynamic cost requirements and / or chassis passability requirements when traveling on the path. In this way, the present application can comprehensively consider the geographical environment of the current driving environment and the current state of the vehicle to perform path planning, ensuring that the target path obtained by path planning meets the dynamic cost requirements and / or chassis passability requirements, thereby achieving reasonable path planning and providing improved passability and safety for unmanned vehicles in off-road environments.

[0010] Additional aspects and advantages of the embodiments of the present application will be given in part in the description below, and in part will become obvious from the description below, or will be learned through practice of the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0012] Figure 1 is a flowchart of a path planning method according to certain embodiments of the present application;

[0013] Figure 2 is a flowchart of a path planning method according to certain embodiments of the present application;

[0014] Figure 3 is a flowchart of a path planning method according to certain embodiments of the present application;

[0015] Figure 4 is a flowchart of a path planning method according to certain embodiments of the present application;

[0016] Figure 5 is a flowchart of a path planning method according to certain embodiments of the present application;

[0017] Figure 6 is a flowchart of a path planning method according to certain embodiments of the present application;

[0018] Figure 7 is a flowchart of a path planning method according to certain embodiments of the present application;

[0019] Figure 8 is a flowchart of a path planning method according to certain embodiments of the present application;

[0020] Figure 9 is a flowchart of a path planning method according to certain embodiments of the present application;

[0021] Figure 10 is a flowchart of a path planning method according to certain embodiments of the present application;

[0022] Figure 11 is a flowchart of a path planning method according to certain embodiments of the present application;

[0023] Figure 12 is a flowchart of a path planning method according to certain embodiments of the present application;

[0024] Figure 13 is a flowchart of a path planning method according to certain embodiments of the present application;

[0025] Figure 14 is a flowchart of a path planning method according to certain embodiments of the present application;

[0026] Figure 15 is a schematic diagram of a module of an electronic device according to some embodiments of the present application;

[0027] Figure 16 is a schematic diagram of a module of a vehicle according to certain embodiments of the present application;

[0028] Figure 17 is a schematic diagram of a connection state between a non-volatile computer-readable storage medium and a processor in certain embodiments of the present application;

[0029] Figure 18 It is a module diagram of a computer program product of certain embodiments of the present application. DETAILED DESCRIPTION

[0030] The embodiments of the present application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the embodiments of the present application, and should not be understood as limiting the embodiments of the present application.

[0031] The present invention provides a path planning method, which is described in detail below:

[0032] Off-road path planning involves designing a safe, efficient, and adaptable path for an unmanned vehicle or robot in complex and variable terrain and road conditions. Off-road terrain is complex and unpredictable. Improper path planning can hinder vehicle operation and even compromise safety. For example, choosing the wrong path can create a risk of rollover when navigating a slope. Therefore, optimal path planning is a pressing issue.

[0033] Existing technologies focus solely on single global or local route planning, lacking comprehensive path planning. In search algorithms, existing technologies only incorporate static stability costs and road surface semantic costs for planning, taking only terrain and road surface information into account. However, if the chassis bottoms out while climbing an off-road vehicle, it will completely lose power and become stuck. If the wrong path is chosen, there is a risk of rollover during the climb. Existing technologies fail to consider these scenarios, neglecting the crucial factors of vehicle passability and dynamic stability in off-road driving.

[0034] To solve the above problems, please refer to Figure 1 The present application provides a path planning method, which is applied to a vehicle and includes:

[0035] Step 01: Path planning is performed based on the vehicle's state information and road information to generate a target path. When the vehicle is traveling along the target path, the vehicle's state information meets preset dynamic cost requirements and / or chassis passability requirements.

[0036] Specifically, road information is information that reflects the geographic environment of the vehicle's current driving environment. For example, road information includes obstacle information, elevation information, and road surface semantics. Appropriate sensors can be used to scan the current driving environment in advance to determine the geographic environment and, therefore, the road information corresponding to the current driving environment. Vehicle status information is information that reflects the vehicle's current driving state, such as vehicle speed or tilt angle. Vehicle status information can be acquired in real time using the vehicle's sensors.

[0037] Dynamic cost refers to the impacts a vehicle experiences while driving on a path, including path smoothness, obstacle collision risk, deviation from the reference path, driving distance, and possible boundary conditions. This dynamic cost reflects whether the vehicle can safely navigate the current driving environment. Chassis passability refers to the ability of the vehicle's chassis to maintain a certain distance from the ground during driving to avoid collisions, scrapes, and sinking into low, uneven surfaces such as mud and sand.

[0038] Dynamic cost requirements and chassis passability requirements can be set based on the vehicle's safe driving state. For example, the dynamic cost requirement can be a low dynamic cost, while the chassis passability requirement can be sufficient. The driving requirements for the target path are then determined based on the driving requirements. For example, the driving requirement can be that the vehicle's state information on the target path meets the dynamic cost requirement, the chassis passability requirement, or both.

[0039] Therefore, when planning the target path corresponding to the current driving environment, planning can be based on the vehicle's status information and road information to simultaneously consider the terrain and topography factors of the current driving environment, and whether the vehicle can safely drive or pass on a certain path.

[0040] At this point, based on the road information and the vehicle's status information, it can first be determined whether the vehicle meets the driving requirements corresponding to the target path when traveling on various paths in the current driving environment, that is, whether it meets the dynamic cost requirements and / or chassis passability requirements. Then, based on the paths that meet the driving requirements, the target path is determined for path planning. It can be understood that the target path is actually the planned path. The vehicle can subsequently travel in the current driving environment according to the target path. In this way, when the vehicle travels on the target path, its status information can meet the dynamic cost requirements and / or chassis passability requirements, so the probability of safety issues occurring when the vehicle travels on the target path is low.

[0041] The path planning method of the embodiments of the present application can, based on road information and vehicle status information, determine a target path that ensures that the vehicle's status information meets preset dynamic cost requirements and / or chassis passability requirements when traveling on the path. In this way, the present application can comprehensively consider the geographical environment of the current driving environment and the current state of the vehicle to perform path planning, ensuring that the target path obtained by path planning meets the dynamic cost requirements and / or chassis passability requirements, thereby achieving reasonable path planning and providing improved passability and safety for unmanned vehicles in off-road environments.

[0042] See also Figure 2 In some embodiments, step 01: performing path planning based on vehicle status information and road information to generate a target path includes:

[0043] Step 011: Plan an initial path based on road information;

[0044] Step 012: Adjust the initial path based on the vehicle's status information to generate a target path.

[0045] Specifically, the geographical environment of the current driving environment can be determined based on the road information of the current driving environment. Based on the geographical environment of the current driving environment, an initial path is planned that satisfies the vehicle's driving requirements. The initial path is the path with the lowest cost in the geographical environment. In this way, the planned initial path is the path with the lowest cost in the geographical environment.

[0046] The initial path described above only considers the terrain and topography of the current driving environment, without considering whether the vehicle can safely navigate or pass through it. Therefore, the initial path should only be used as a reference. Further consideration should be given to the vehicle's dynamic cost and chassis maneuverability when planning a local path.

[0047] At this point, the initial path can also be adjusted to generate a target path, ensuring that the vehicle's status information meets the dynamic cost requirements and / or chassis passability requirements when traveling on the target path. For example, a candidate path can be searched near the initial path. Then, based on the road information of the candidate path and the vehicle's current status information, a determination is made as to whether the vehicle can meet the dynamic cost requirements and chassis passability requirements when traveling on the candidate path. If the vehicle's status information meets the preset dynamic cost requirements and chassis passability requirements when traveling on a candidate path, the candidate path is selected as the target path. As can be understood, the probability of safety issues occurring when subsequent vehicles travel on the target path is reduced.

[0048] In this way, an initial path that meets the vehicle's driving needs based on the geographical environment can be obtained first based on road information, and then adjustments can be made on the basis of the initial path to obtain a target path that ensures that when the vehicle is driving on the path, the vehicle's status information meets the preset dynamic cost requirements and / or chassis passability requirements, thereby achieving reasonable path planning and providing better passability and safety for unmanned vehicles in off-road environments.

[0049] See also Figure 3 In some embodiments, step 012: adjusting the initial path based on the vehicle state information to generate a target path includes:

[0050] Step 0121: Based on the current path point and the initial path, obtain multiple candidate paths corresponding to the current path point;

[0051] Step 0122: When it is determined that the vehicle's status information meets the dynamic cost requirement and / or chassis passability requirement, the corresponding candidate path is the target path.

[0052] Specifically, the current path point is the path point the vehicle is currently located at in the current driving environment. The initial path includes a sequence of movement between multiple path points. The path between the current path point and the next path point to be moved to is the planned path corresponding to the current path point. Multiple candidate paths can then be identified near the planned path. It will be understood that the planned path conforms to the geographical environment of the current driving environment, and therefore, candidate paths near the planned path can also conform to, or approximately conform to, the geographical environment of the current driving environment.

[0053] Next, the system determines whether the vehicle, in its current state, can meet the dynamic cost requirements and / or chassis passability requirements when traveling along each candidate path. For example, a mathematical model can be constructed between the vehicle's state information, the road information of the candidate path, and the dynamic cost requirements, and a mathematical model can be constructed between the vehicle's state information, the road information of the candidate path, and the chassis passability requirements. The vehicle's state information and the road information of the candidate path are then incorporated into these two mathematical models to determine whether the vehicle meets the dynamic cost requirements and / or chassis passability requirements when traveling along each candidate path.

[0054] Next, the target path is determined to be the candidate path corresponding to the vehicle's state information meeting the dynamic cost requirements and / or chassis passability requirements. It is understood that the target path is actually the local path that the vehicle needs to take next. Therefore, the target path is actually a local path that is adjusted in real time, not a global path.

[0055] In this way, the path that the vehicle needs to take next in the initial path can be adjusted in real time, and multiple candidate paths can be determined near the planned path corresponding to the initial path. The path planning is carried out by integrating terrain, road surface information, vehicle passability, and dynamic stability constraints to obtain a target path that can ensure that when the vehicle travels on the path, the vehicle's status information meets the preset dynamic cost requirements and / or chassis passability requirements, thereby achieving reasonable path planning and providing better passability and safety for unmanned vehicles in off-road environments.

[0056] See also Figure 4 In some embodiments, step 0121: based on the current path point and the initial path, obtaining multiple candidate paths corresponding to the current path point includes:

[0057] Step 01211: Determine the moving path point corresponding to the current path point in the initial path;

[0058] Step 01212: Determine multiple sampling points corresponding to the moving path point based on a preset sampling rule;

[0059] Step 01213: Determine multiple candidate paths corresponding to the current path point based on multiple sampling points of the current path point and the moving path point.

[0060] Specifically, the next pathpoint to which the vehicle needs to move, i.e., the moving pathpoint, can be determined based on the movement order of the pathpoints in the initial path. Then, multiple sampling points are acquired near the moving pathpoint based on a preset sampling rule. The preset sampling rule is a sampling rule that allows for the selection of sampling points near the moving pathpoint whose geographical environment is not significantly different from that of the moving pathpoint.

[0061] In certain embodiments, the number of sampling points is determined based on the distance between the current path point and the moving path point and a preset offset. For example, the distance between the current path point and the moving path point is divided by the preset offset to obtain the number of sampling points. Then, based on the number of sampling points, sampling points of the moving path point are determined on a line perpendicular to the line between the current path point and the moving path point. This sampling point can be understood as a lateral sampling point of the moving path point. Because the distance between the current path point and the moving path point varies, the number of sampling points is dynamically adjusted. When the distance is large, the number of sampling points is increased; otherwise, the number of sampling points is reduced, thereby avoiding the generation of candidate paths with large bending energy.

[0062] Next, multiple candidate paths corresponding to the current path point can be determined based on the multiple sampling points of the current path point and the moving path point. For example, a k-order B-spline algorithm can be used to fit the multiple sampling points of the current path point and the moving path point. Preferably, a cubic B-spline algorithm can be used to generate multiple candidate paths.

[0063] In this way, the moving path point of the current path point can be obtained based on the initial path, and then a sampling point whose geographical environment is not much different from the geographical environment of the moving path point can be found near the moving path point based on the preset sampling rules. Finally, the selected path is determined based on the sampling point and the current path point, so that the geographical environment of the selected path and the planned path in the initial path is not much different, ensuring that the geographical environment of the target path subsequently obtained based on the selected path meets the driving needs of the vehicle.

[0064] See also Figure 5 In some embodiments, step 0122: when it is determined that the vehicle state information meets the dynamic cost requirement and / or the chassis passability requirement, the corresponding candidate path is the target path, including:

[0065] Step 01221: Determine the dynamic cost of the vehicle when it is located on each candidate path and / or the availability of the chassis passability based on the vehicle status information;

[0066] Step 01222: Determine whether the dynamic cost of the vehicle meets the dynamic cost requirement, and / or whether the chassis passability of the vehicle meets the chassis passability requirement. The candidate path is the target path.

[0067] Specifically, dynamic cost refers to the impact a vehicle experiences while traveling on a potential path. For example, dynamic cost includes dynamic stability cost. Dynamic stability refers to the vehicle's balance and handling capabilities during driving. This stability is directly related to driving safety and is a key performance indicator in automotive design and manufacturing. Specifically, a vehicle may be subject to various external disturbances during driving, such as crosswinds, uneven roads, and sharp turns. A vehicle with good dynamic stability can quickly recover its original driving state and direction under these disturbances, preventing loss of control, skidding, or rollover.

[0068] The availability of chassis passability refers to whether the chassis has passability.

[0069] The vehicle's dynamic cost and / or chassis passability are determined based on the vehicle's state information. For example, a mathematical model can be constructed between the vehicle's state information and the dynamic cost, or a mathematical model can be constructed between the vehicle's state information and the chassis passability. The vehicle's state information is then incorporated into the corresponding mathematical model to determine the vehicle's dynamic cost and chassis passability when traveling on the selected path.

[0070] Then, the target path can be determined as the candidate path whose dynamic cost satisfies the dynamic cost requirement and / or whose chassis passability meets the chassis passability requirement. Specifically, the target path can be determined as the candidate path whose dynamic cost satisfies the dynamic cost requirement, or the target path whose chassis passability meets the chassis passability requirement, or the target path whose dynamic cost satisfies the dynamic cost requirement and whose chassis passability meets the chassis passability requirement. The specific determination can be made based on actual needs.

[0071] For example, if the dynamic cost requirement includes a minimum dynamic cost and the chassis passability requirement includes that the vehicle has chassis passability, then the candidate path with the minimum dynamic cost and chassis passability can be determined as the target path.

[0072] In this way, the dynamic cost of the selected path and the chassis passability of the selected path can be accurately judged based on the vehicle's status information, so as to accurately obtain the target path in which the dynamic cost meets the dynamic cost requirements and / or the chassis passability meets the chassis passability requirements, thereby ensuring that the vehicle can safely travel on the target path.

[0073] See also Figure 6In some embodiments, step 01221: determining the dynamic cost and / or chassis passability of the vehicle based on the vehicle status information includes:

[0074] Step 012211: Determine the dynamic cost of the vehicle when it is on the candidate path based on the current state information of the vehicle and the road information of the candidate path; and / or

[0075] Step 012212: Based on the current state information of the vehicle and the road information of the path to be selected, determine the chassis passability of the vehicle when it is on the path to be selected.

[0076] Specifically, in some embodiments, parameters that affect the dynamic cost can be determined from the vehicle's current state information and the road information of the candidate path. Based on these parameters, a mathematical model is constructed between the vehicle's current state information, the road information of the candidate path, and the dynamic cost. The vehicle's current state information and the road information of the candidate path are then incorporated into this mathematical model to determine the dynamic cost for the vehicle when it is on the candidate path.

[0077] In certain embodiments, parameters that affect the chassis's passability can be determined from the vehicle's current state information and the road information of the candidate route. Based on these parameters, a mathematical model is constructed that relates the vehicle's current state information, the road information of the candidate route, and the chassis's passability. The vehicle's current state information and the road information of the candidate route are then incorporated into the mathematical model to determine the chassis's passability when the vehicle is on the candidate route.

[0078] In this way, the dynamic cost of the vehicle and the chassis passability can be accurately determined based on the vehicle's current state information and the road information of the selected path.

[0079] See also Figure 7 In some embodiments, step 012211: determining a dynamic cost when the vehicle is located on the candidate path based on the current state information of the vehicle and the road information of the candidate path, includes:

[0080] Step 012213: Determine the road surface cost of the candidate path based on the road information of the candidate path;

[0081] Step 012214: Determine the vehicle's operating cost based on the vehicle's current state information and the road information of the selected path;

[0082] Step 012215: Determine the dynamic cost based on the road surface cost and the operation cost.

[0083] Specifically, road cost measures the various road-related costs and impacts a vehicle faces while traveling along a particular route. Parameters within the road information of a candidate route that influence the road cost are identified. Based on these parameters, a mathematical model is constructed that relates the road information of the candidate route to the road cost. The road information of the candidate route is then incorporated into this mathematical model to determine the road cost for the vehicle on the candidate route.

[0084] The operating cost measures the various costs and impacts a vehicle faces while traveling on a route. Parameters that influence the operating cost are determined from the vehicle's current state and the road information of the candidate route. Based on these parameters, a mathematical model is constructed that relates the vehicle's current state, the road information of the candidate route, and the operating cost. The current state and road information of the candidate route are then incorporated into this mathematical model to determine the operating cost for the vehicle on the candidate route.

[0085] Then, the dynamic cost may be determined based on the road surface cost and the operation cost, for example, based on the road surface cost, a weight coefficient corresponding to the road surface cost, the operation cost, and a weight coefficient corresponding to the operation cost.

[0086] In this way, the current state information of the vehicle and the road information of the selected path can be combined to accurately determine the road surface cost and operating cost of the vehicle on the selected path, thereby accurately obtaining the dynamic cost of the vehicle on the selected path, so as to facilitate the subsequent acquisition of the target path based on the dynamic cost.

[0087] In some embodiments, the road surface cost includes at least one of a target distance between a sampling point of the candidate path and the current path point and a road surface semantic cost.

[0088] Specifically, the target distance between the sampling point of the candidate path and the current path point can be the Euclidean distance between the end point and the starting point of the candidate path, that is, the Euclidean distance between the current path point and the sampling point corresponding to the candidate path. Euclidean distance is a method for calculating the straight-line distance between two points in n-dimensional space. Its calculation formula is: Among them, (x1, y1, z1) and (x2, y2, z2) represent two points in space respectively. By calculating the Euclidean distance of the vehicle from the starting point to the end point, the scope and degree of vehicle movement can be intuitively understood, thereby more accurately evaluating the road cost. The Euclidean distance between the current path point and the candidate path can be calculated based on the coordinates of the current path point and the coordinates of the candidate path in the terrain elevation map of the current driving environment. Among them, the terrain elevation map is a map that shows the ups and downs of the surface. The terrain elevation map establishes a coordinate system based on the current driving environment, and any position point in the current driving environment has corresponding coordinates in the terrain elevation map.

[0089] The road surface semantic cost refers to the semantically related cost or price faced by an autonomous vehicle when traveling on a specific road surface during the path planning process. Based on road information, the road surface types included in the candidate path can be determined, as well as the proportion of different road surface types in the candidate path. For example, road information includes a road surface semantic map corresponding to the current driving environment. A road surface semantic map is a map representation that contains road geometry and semantic information, and includes various road surface types in the current driving environment. Based on the road surface semantic map, the proportion of different road surface types in the candidate path can be determined. The road surface semantic cost is then determined based on the proportion of different road surface types in the candidate path and the weight coefficients corresponding to each road surface type.

[0090] Semantic cost of road surface i ) is:

[0091]

[0092] Among them, P j (n i ) represents the proportion of different road types at the i-th child node of the selected path, and road types include but are not limited to dirt road, grass, sand, snow, etc., ω j It represents the weight coefficient corresponding to different road types. The larger the weight coefficient, the less likely it is that the vehicle is to travel on this road type.

[0093] In certain embodiments, the operating cost includes at least one of a dynamic stability cost and a static stability cost of the vehicle.

[0094] Specifically, in some embodiments, when calculating dynamic stability, the dynamic stability of the selected path can be determined based on the lateral acceleration corresponding to the selected path. In this case, the selected path can be divided into multiple nodes, for example, the selected path can be evenly divided into k nodes. The current state information includes the current speed v of the vehicle, and the road information includes the curvature radius R corresponding to each node on the selected path. j (j=1,2,...k). Then, according to the current speed v and the curvature radius R corresponding to each node j Determine the lateral acceleration corresponding to each node. The calculation formula is:

[0095]

[0096] Among them, a yj is the lateral acceleration of the vehicle at the jth node.

[0097] Then, the dynamic stability cost is determined based on the lateral acceleration corresponding to each node. For example, the dynamic stability cost on the rth candidate path can be obtained by averaging the lateral acceleration corresponding to all nodes. The calculation formula in this case is:

[0098]

[0099] Among them, Stable_Dyna(l r ) is the dynamic stability cost of the vehicle on the rth candidate path.

[0100] Static stability refers to the stability of a vehicle when parked or at rest. This is primarily determined by analyzing static equilibrium conditions. A vehicle's stability at rest depends primarily on factors such as its mass distribution, suspension system, and tire-ground contact. A vehicle with good static stability maintains balance when parked or at rest, and is less likely to tip over or slide.

[0101] In some embodiments, the road information includes the pitch angle and roll angle of the current path point, as well as the pitch angle and roll angle of the sampling point corresponding to the selected path. When calculating static stability, the static stability cost is determined based on the pitch angle and roll angle of the current path point, as well as the pitch angle and roll angle of the sampling point corresponding to the selected path. In this case, the static stability cost represents the slope information of the terrain. The pitch angle and roll angle of the current path point parent, as well as the pitch angle and roll angle of the sampling point corresponding to the selected path, are the static stability cost. i The pitch angle and roll angle of the corresponding sampling point can be obtained based on the terrain elevation map of the current driving environment. The calculation formula is:

[0102]

[0103] where Δθ n-parent (n i ) represents the absolute value of the pitch angle difference between the current path point and the sampling point, Δφ n-parent (n i ) represents the absolute value of the roll angle difference between the current path point and the sampling point.

[0104] In some embodiments, the dynamic cost may be determined based on the road cost, the weight coefficient corresponding to the road cost, the operating cost, and the weight coefficient corresponding to the operating cost. For example, the dynamic cost may be calculated according to the following formula:

[0105] f(l r )=ω d d(l r )+ω dy Stable_Dyna(l r )+ω st Stable(l r)+ω se Semantic(l r )

[0106] Among them, d(l r ) represents the distance between the sampling point of the selected path and the current path point, Stable_Dyna(l r ) represents the dynamic stability cost, Stable(n i ) represents the static stability cost, Semantic(n i ) represents the road semantic cost, ω st is the weight coefficient of static stability cost, ω se is the weight coefficient of the road semantic cost, ω dy is the weight coefficient of dynamic stability cost, ω d is the weight coefficient of the distance between the sampling point of the path to be selected and the current path point.

[0107] See also Figure 8 In some embodiments, step 012212: determining the chassis passability of the vehicle when it is on the to-be-selected path based on the current state information of the vehicle and the road information of the to-be-selected path, includes:

[0108] Step 012216: Substitute the current state information and the road information of the selected path into a preset passability equation to obtain a passability parameter, where the passability parameter includes the height difference between the ground of the selected path and the chassis of the vehicle;

[0109] Step 012217: When the passability parameters meet the preset passability conditions, it is determined that the vehicle has chassis passability when it is located on the to-be-selected path.

[0110] Specifically, parameters that may affect the area where the vehicle's chassis is located in the vehicle's state information may be predetermined, and then a first plane equation may be constructed based on the parameters to determine the vehicle's state information and the position of the area where the vehicle's chassis is located in the current driving environment. For example, the state parameters include the vehicle's four-wheel suspension height, the vehicle's position, and one of the vehicle's body pitch angle and roll angle. Parameters that may affect the position of the selected path in the current driving environment may be predetermined in the road information of the selected path, and then a second plane equation may be constructed based on the parameters to determine the road information of the selected path and the position of the selected path in the current driving environment. For example, the road information includes the coordinates of the selected path in the current driving environment. It will be understood that the heights calculated by these two plane equations are based on the zero point of the coordinate system of the current driving environment, that is, absolute heights, not relative heights. The position of an object in the current driving environment may be obtained based on the coordinates of the object in the terrain elevation map of the current driving environment.

[0111] The difference between the second plane equation and the first plane equation represents the height difference between the ground surface of the selected path and the vehicle's chassis. Therefore, by subtracting the first plane equation from the second plane equation, a preset passability equation representing the height difference between the ground surface of the selected path and the vehicle's chassis, when the vehicle is located on different selected paths and in different vehicle states, can be obtained.

[0112] Therefore, the current state information and the road information of the selected path can be substituted into a preset passability equation to obtain a passability parameter. The passability parameter includes the height difference between the ground surface of the selected path and the vehicle's chassis. If the passability parameter satisfies the preset passability condition, the vehicle has chassis passability while on the selected path. Furthermore, in certain embodiments, if the passability parameter does not satisfy the preset passability condition, the vehicle does not have chassis passability while on the selected path.

[0113] In one embodiment, the preset passability condition includes a passability parameter being less than a preset value. For example, the preset value is 0. If the currently calculated passability parameter is less than 0, it is considered that the vehicle's chassis is above the road surface of the selected path. In this case, the vehicle is considered to have chassis passability and can safely pass through the selected path. If the passability parameter is greater than 0, it is considered that the vehicle's chassis is below the road surface of the selected path and cannot safely pass through the selected path. In this case, the vehicle is considered to have no chassis passability. As another example, the preset value is a negative safety distance. The safety distance is the distance between the vehicle's chassis and the road surface of the selected path that ensures safe passage of the selected path. If the currently calculated passability parameter is less than the negative safety distance, it is considered that the vehicle's chassis is above the road surface of the selected path and the distance between the two is greater than the safety distance. In this case, the vehicle is considered to have chassis passability and can safely pass through the selected path. If the passability parameter is greater than the safety distance, it is considered that the distance between the vehicle's chassis and the road surface of the selected path is less than the safety distance. The vehicle may not be able to safely pass through the selected path and the vehicle is considered to have no chassis passability.

[0114] In some embodiments, the candidate path may be divided into multiple nodes, and then the passability parameter corresponding to each node is calculated, and then whether the vehicle has chassis passability is determined based on the calculation result of each node.

[0115] The state information and coordinates of each node on the candidate path are then substituted into a preset passability equation to determine the passability parameters for each node. If the passability parameters for all nodes on the candidate path meet the preset passability conditions, the vehicle is considered to have chassis passability.

[0116] For example, the chassis of the vehicle can be determined based on the height of the four-wheel suspension, the position of the vehicle, and the pitch angle and roll angle of the vehicle body, or the preset passability equation S(n) of the bottom plate of the battery pack can be directly constructed, and the three-dimensional coordinates n of the k nodes on the selected path can be sequentially calculated. j (x j ,y j ,z j ) is substituted into the preset passability equation. If any node satisfies:

[0117] S(n j )>0(j=1…k)

[0118] The candidate path is then deemed to not meet the vehicle passability constraints, meaning it lacks chassis passability. This indicates that there is a node on the path above the vehicle chassis, which could cause the battery pack's lower floor to bottom out. This bottoming out of the off-road vehicle will cause it to lose power and become stuck, so this situation should be avoided as much as possible. Specifically, when the target road section is a section where the vehicle's status information meets the dynamic cost and chassis passability requirements, in order to improve the search efficiency when searching for paths that meet these requirements, all paths that do not meet the vehicle passability constraints, i.e., paths that lack chassis passability, should be eliminated before the search begins.

[0119] In this way, the preset passability equation can be used to calculate in real time whether the vehicle meets the vehicle passability constraints when driving on the selected path, that is, whether the chassis passability is met, thereby improving the accuracy of subsequent target path selection.

[0120] In summary, this application can integrate terrain, road surface information with vehicle passability and dynamic stability constraints for path planning, providing better passability and safety for unmanned vehicles in off-road environments.

[0121] See also Figure 9 In some embodiments, step 011: planning an initial path based on road information includes:

[0122] Step 0111: Determine the planned partial path corresponding to the current planned path point based on the current planned path point and road information. The current planned path point at least includes the starting point of the current driving environment. The static cost corresponding to the planned partial path meets the static cost requirement.

[0123] Step 0112: The end point of the planned partial path is used as the next current planned path point, and the process again proceeds to the step of determining the planned partial path corresponding to the current planned path point based on the current planned path point and road information, until the planned partial path includes the end point of the trip in the current driving environment;

[0124] Step 0113: Determine the initial path based on all planned local paths of the current driving environment.

[0125] Specifically, the currently planned path point is the starting point of the current driving environment, or the end point of the planned partial path corresponding to any path point in the current driving environment. The starting point of the current driving environment refers to the starting point of the vehicle's entire journey within the current driving environment. This means that the initial path is planned segment by segment.

[0126] First, the starting point of the current travel environment is used as the current planned path point, and a planned local path for the current planned path point is determined. Based on the current planned path point and road information, the static cost of the paths near the current planned path point can be determined. The static cost represents the additional cost or impact incurred by the vehicle during travel due to relatively stable factors such as the current travel environment's geographical environment, such as topography or landforms. A mathematical model can be pre-established between the road information and the static cost. The road information corresponding to the current planned path point is then incorporated into this mathematical model to determine the static cost of the paths near the current planned path point. A static cost requirement can then be set based on the static cost of each path at the highest safety level. For example, the static cost requirement can be set such that the static cost corresponding to the planned local path is the minimum static cost among the paths near the current planned path point. The paths near the current planned path point whose static costs meet the static cost requirement are then determined as the planned local path corresponding to the current planned path point. For example, the path with the minimum static cost is determined as the planned local path.

[0127] The end point of the planned local path is then used as the next current planned path point, and step 0111 is re-entered. This cycle repeats until a planned local path includes the end point of the current driving environment, thus completing the global path planning for the current driving environment. The end point of the current driving environment refers to the end point of the vehicle's entire journey in the current driving environment.

[0128] Finally, all planned local paths are connected in the order of their starting points to obtain the initial path. It can be understood that the initial path is the global path of the current driving environment.

[0129] In this way, the static cost of the path near the current planned path point can be calculated first based on the road information of the current driving environment to determine the planned local path whose corresponding static cost can meet the static cost requirements, so as to obtain the initial path whose geographical environment can meet the driving needs of the vehicle based on the planned local path.

[0130] See also Figure 10 In some embodiments, step 0111: determining a planned local path corresponding to the current planned path point based on the current planned path point and road information, includes:

[0131] Step 01111: Determine multiple candidate path points corresponding to the current planned path point based on the current planned path point and road information;

[0132] Step 01112: Determine the static cost of each candidate path point based on the road information corresponding to each candidate path point;

[0133] Step 01113: Based on the candidate path points whose static costs meet the static cost requirements, determine the selected path points of the current planned path point to determine the planned local path of the current planned path point.

[0134] Specifically, first, based on the currently planned pathpoint and road information, multiple candidate pathpoints that the vehicle can next travel to within the area near the currently planned pathpoint can be determined. For example, obstacles in the area near the currently planned pathpoint can be obtained based on road information, and then multiple candidate pathpoints that the vehicle can next travel to can be determined based on the obstacles.

[0135] Next, the static cost of each candidate pathpoint is determined based on the road information corresponding to each candidate pathpoint. The static cost represents the additional cost or impact incurred by the vehicle during driving due to relatively stable factors such as the current driving environment, such as topography or landforms. A mathematical model linking road information and static cost can be pre-built. The static cost corresponding to each candidate pathpoint can then be used to determine the impact on the vehicle after reaching each candidate pathpoint.

[0136] Then, a static cost requirement can be set based on the static cost when the vehicle is safer after traveling to each candidate path point, for example, the static cost requirement is to minimize the static cost. Then, a candidate path point whose static cost meets the static cost requirement is searched among the candidate path points, for example, by using the A-star algorithm to search, and the candidate path point with the smallest static cost is selected in each search step. The selected path point of the current planned path point is determined, and the path between the current planned path point and the selected path point is the planned local path of the current planned path point. The static cost corresponding to the planned local path described above is actually the static cost of the selected path point corresponding to the planned local path.

[0137] Finally, all planned local paths are connected in the order of their starting points to obtain the initial path. It can be understood that the initial path is the global path of the current driving environment.

[0138] In this way, the static cost of the candidate path points corresponding to the current planned path point can be calculated first based on the road information of the current driving environment, and then the planned local path corresponding to the current planned path point can be determined based on the cost of each candidate path point, so as to obtain the initial path whose geographical environment can meet the driving needs of the vehicle according to the planned local path.

[0139] See also Figure 11 In some embodiments, the road information includes an obstacle map and a terrain elevation map of the current driving environment; Step 01111: determining multiple candidate path points corresponding to the current planned path point based on the current planned path point and the road information, including:

[0140] Step 011111: Based on the current position of the currently planned path point and the obstacle map, obtain the target moving distance;

[0141] Step 011112: Determine multiple initial path points corresponding to the current planned path point based on the target moving distance, the current planned path point, and the preset angle;

[0142] Step 011113: Determine multiple candidate path points corresponding to the current planned path point based on the target moving distance, initial path points and terrain elevation map.

[0143] Specifically, the obstacle map contains the location information of each obstacle in the current driving environment. Based on the obstacle map and the current location of the currently planned pathpoint, the obstacles near the currently planned pathpoint can be determined. Then, based on the obstacles near the currently planned pathpoint and the distance between the obstacles and the currently planned pathpoint, the target movement distance can be determined.

[0144] For example, based on the obstacle map, the closest distance between the current location of the currently planned path point and an obstacle can be obtained, with this closest distance serving as the initial movement distance. In this case, the location of any obstacles near the currently planned path point can be first obtained, followed by the closest distance between the current location of the currently planned path point and the obstacle, with this closest distance serving as the initial movement distance. It will be understood that the initial movement distance is the closest distance the vehicle can travel when moving near the currently planned path, as determined by the obstacle map.

[0145] Then, the target moving distance is determined based on the initial moving distance and the preset distance range. The preset distance range is a reasonable range of distances for the vehicle's next travel. If the preset distance range is too short, there will be too many candidate paths, which will affect path planning. If the preset distance range is too long, the path planning accuracy will be low. When the initial moving distance is within the preset distance range, the initial moving distance is used as the target moving distance; when the initial moving distance is less than the minimum distance of the preset distance range, the minimum distance is taken as the target moving distance; when the initial moving distance is greater than the maximum distance of the preset distance range, the maximum distance is taken as the target moving distance, that is:

[0146]

[0147] where d max Indicates the target moving distance, R indicates the initial moving distance of the current planned path point from the obstacle,

[0148] [R min ,R max ] indicates the preset distance range.

[0149] Next, a path circle is generated with the current planned path point as the center and the target movement distance as the radius. Initial path points are then acquired on the path circle at preset angles to obtain multiple initial path points corresponding to the current planned path point. The obstacle map is only a two-dimensional map, so the target movement distance is two-dimensional data. However, the current driving environment is three-dimensional. Therefore, in real-world environments, the distance between the initial path point and the current planned path point may not accurately represent the distance the vehicle can travel next.

[0150] Then, based on the target movement distance, the initial pathpoint, and the terrain elevation map, the initial pathpoint can be adjusted based on the altitude information in the terrain elevation map. For example, the distance between the current planned pathpoint and the candidate pathpoint can be determined based on the initial pathpoint and the altitude information, and then multiple candidate pathpoints can be determined based on this distance. In this way, the altitude factor can be added to determine the multiple candidate pathpoints corresponding to the current planned pathpoint. It can be understood that the candidate pathpoints are three-dimensional pathpoints.

[0151] In this way, based on road information, the candidate path points that the vehicle can actually move to next when it is located at the current planned path point can be accurately obtained, thereby improving the accuracy of subsequent determination of the planned local path.

[0152] See also Figure 12 In some embodiments, step 011111: obtaining a target moving distance based on a current position of a currently planned path point and an obstacle map, includes:

[0153] Step 011114: Based on the target initial path point and the terrain elevation map, obtain the intermediate path point corresponding to the target initial path point on the terrain elevation map, where the target initial path point is any initial path point;

[0154] Step 011115: Based on the target moving distance, the height of the intermediate path point in the terrain elevation map and the height of the current planned path point in the terrain elevation map, determine the candidate path points corresponding to the intermediate path point to determine multiple candidate path points corresponding to the current planned path point.

[0155] Specifically, first, the target initial path point is projected into the terrain elevation map to obtain the intermediate path point corresponding to the target initial path point in the terrain elevation map. At this time, the intermediate path point is a three-dimensional path point, and the target initial path point is any initial path point.

[0156] In one embodiment, the unit climbing height between the intermediate path point and the current planned path point is obtained based on the height of the intermediate path point on the terrain elevation map and the height of the current planned path point on the terrain elevation map. Then, the dynamic movement distance of the current planned path point in the direction of the intermediate path point is obtained by combining the unit climbing height and the target movement distance of the obstacle layer. The direction of the intermediate path point is the direction of the current planned path point toward the intermediate path point. The dynamic movement distance is calculated as follows:

[0157] The coordinates of the current planned path point are represented as n parent (x parent ,y parent ,z parent ), the coordinates of the jth intermediate path point are expressed as n j (x j ,y j ,z j ), the unit climbing height Δz(n j ) is:

[0158]

[0159] This allows us to calculate the dynamic moving distance of the current planned path point in each three-dimensional sub-direction:

[0160]

[0161] where d j Indicates the distance from the current planned path point to the jth intermediate path point, that is, the dynamic moving distance, κ max Indicates the preset maximum curvature value.

[0162] Finally, based on the current planned pathpoint, the dynamic movement distance, and the direction of the intermediate pathpoint, the candidate pathpoint corresponding to the intermediate pathpoint is determined. That is, among the points along the direction of the intermediate pathpoint, the point whose distance from the current planned pathpoint is the dynamic movement distance is selected as the candidate pathpoint corresponding to the intermediate pathpoint, thereby updating the candidate pathpoint corresponding to the current planned pathpoint. Pathpoints can then be selected based on the search among the candidate pathpoints.

[0163] In this way, the candidate path points can be updated by combining the target moving distance and the information of the terrain elevation map, thereby ensuring that the actual distance between the obtained candidate path points and the currently planned path points is reasonable.

[0164] See also Figure 13 In some embodiments, step 01112: determining the static cost of each candidate path point based on the road information corresponding to each candidate path point includes:

[0165] Step 011121: Determine the movement cost of the target candidate path point based on the road information corresponding to the current planned path point and the target candidate path point, where the target candidate path point is any candidate path point;

[0166] Step 011122: Determine the heuristic cost of the target candidate pathpoint based on the road information corresponding to the target candidate pathpoint and the road information corresponding to the trip end point of the current driving environment;

[0167] Step 011123: Determine the static cost of the target candidate path point based on the movement cost and the heuristic cost.

[0168] Specifically, the movement cost represents the impact of the terrain and topography on the vehicle as it moves from the current planned pathpoint to the target candidate pathpoint. The heuristic cost represents the impact of the terrain and topography on the vehicle as it moves from the target candidate pathpoint to the end of the current driving environment. The current planned pathpoint is either the starting point of the current driving environment or the target candidate pathpoint corresponding to a planned path segment. Therefore, the static cost of the current planned pathpoint is known.

[0169] First, a mathematical model can be constructed in advance based on the road information corresponding to the current planned path point, the road information corresponding to the target candidate path point, and the movement cost. The road information corresponding to the current planned path point and the road information corresponding to the target candidate path point are then fed into this mathematical model to determine the movement cost of the target candidate path.

[0170] For example, the movement cost of the target candidate path point includes at least one of the movement cost of the current planned path point, the Euclidean distance between the current planned path point and the target candidate path point, the static stability cost, and the road semantic cost. The current planned path point is the starting point of the current driving environment or the target candidate path point corresponding to a certain planned path segment. Therefore, the movement cost of the current planned path point is known. The expression of the movement cost is:

[0171] g(n i )=g(n parent )+ω d d(n i )+ω st Stable(n i )+ω se Semantic(n i )

[0172] Where d(n i ) represents the current planned path point n parent To the target candidate path point n i Euclidean distance, Stable(n i ) represents the static stability cost between the current planning path point and the target candidate path point, Semantic(n i ) represents the road semantic cost between the current planned path point and the target candidate path point. d 、ω st 、ω se Expressed as the weight coefficient of the corresponding cost.

[0173] Euclidean distance is a method for calculating the straight-line distance between two points in n-dimensional space. Its calculation formula is: Where (x1, y1, z1) and (x2, y2, z2) represent two points in space. By calculating the Euclidean distance from the vehicle's starting point to the end point, we can intuitively understand the scope and extent of the vehicle's movement, thereby more accurately evaluating the movement cost.

[0174] The static stability cost can be determined based on the pitch angle of the currently planned pathpoint compared to the pitch angle of the target candidate pathpoint, and / or the roll angle of the currently planned pathpoint compared to the roll angle of the target candidate pathpoint. In this way, the static stability cost can be used to represent the slope of the terrain. It can be understood that the greater the slope, the greater the impact on the vehicle, and the greater the risk of rollover or failure to successfully climb the slope. The pitch and roll angles of both pathpoints can be obtained from a terrain elevation map of the current driving environment.

[0175] Taking the static stability cost as an example, which can be determined based on the pitch angle of the current planned path point and the pitch angle of the target candidate path point, and the roll angle of the current planned path point and the roll angle of the target candidate path point, the calculation formula is:

[0176]

[0177] where Δθ n-parent (n i ) represents the absolute value of the pitch angle difference between the current planned path point and the target candidate path point, Δφ n-parent (n i ) represents the absolute value of the roll angle difference between the current planned path point and the target candidate path point, where parent refers to the current planned path point and n refers to the target candidate path point.

[0178] The road semantic cost comprehensively considers various road surface types and their characteristics, such as dirt roads, grass, and shrubs, and sets different costs for vehicle movement based on these characteristics. In some embodiments, the road information includes a road semantic map corresponding to the current driving environment. Based on the road semantic map, the proportion of different road surface types in the path between the target candidate pathpoint and the currently planned pathpoint can be determined. The road semantic cost is then determined based on the proportion of different road surface types in the path between the target candidate pathpoint and the currently planned pathpoint and the weight coefficients corresponding to each road surface type.

[0179] Semantic cost of road surface i ) is:

[0180]

[0181] Among them, P j (n i ) represents the proportion of different road types at the i-th child node between the current planned path point and the target candidate path point. Road types include but are not limited to dirt roads, grass, sand, snow, etc., ω j It represents the weight coefficient corresponding to different road types. The larger the weight coefficient, the less likely it is that the vehicle is to travel on this road type.

[0182] In this way, the above formula can accurately calculate the movement cost that can accurately consider the impact of the ground topography and terrain on the vehicle when moving from the current planned path point to the target candidate path point.

[0183] A mathematical model can then be constructed in advance, combining the road information corresponding to the current travel environment's end point, the road information corresponding to the target candidate pathpoint, and the movement cost. This mathematical model is then incorporated into the road information corresponding to the current travel environment's end point and the target candidate pathpoint to determine the heuristic cost of the target candidate path. For example, the heuristic cost may include the Euclidean distance between the current travel environment's end point and the target candidate pathpoint.

[0184] Next, the static cost is determined based on the heuristic cost and the movement cost. For example, the sum of the heuristic cost and the movement cost is used as the static cost, which is expressed as:

[0185] f(n i )=g(n i )+h(n i )

[0186] Among them, f(n i ) is the static cost, g(n i ) is the moving cost, h(n i ) is the heuristic cost.

[0187] Each candidate path point is calculated based on the above formula to obtain the static cost corresponding to each candidate path.

[0188] In this way, the above formula can accurately calculate the static cost that can accurately consider the impact of the ground topography and terrain on the vehicle when moving from the current planned path point to the candidate path point, and from the candidate path point to the end of the trip in the current driving environment, thereby facilitating the acquisition of the optimal planned local path based on the static cost of the candidate path point.

[0189] See also Figure 14 In some embodiments, the path planning method further includes:

[0190] Step 02: Determine an environment map corresponding to the current driving environment based on preset image data and preset point cloud data of the current driving environment, where the environment map includes road information.

[0191] Specifically, image and point cloud data can be collected in advance within the current driving environment. For example, images can be captured within the current driving environment and then scanned using radar to generate preset image data and point cloud data corresponding to the current driving environment. The preset image data can generate two-dimensional data of the current driving environment, including objects within the current driving environment and the geographical environment of each area. Point cloud data can provide richer three-dimensional spatial information, such as the height of each object and the distance between them.

[0192] Therefore, road information can be generated based on the preset image data and the preset point cloud data, and an environmental map corresponding to the current driving environment can be constructed based on the road information. It can be understood that the environmental map covers all road information of the current driving environment, such as the terrain and landforms of different areas of the current driving environment, and obstacles in the current driving environment.

[0193] In this way, an environmental map can be constructed in advance based on the preset image data and preset point cloud data of the current driving environment, thereby facilitating the subsequent accurate acquisition of road information based on the environmental map.

[0194] In some embodiments, the environment map includes at least one or more of an obstacle map, a terrain elevation map, and a road semantic map.

[0195] Specifically, the environment map can be a multi-layered map to ensure it accurately reflects the various road information in the current driving environment. In this case, the environment map can be divided into an obstacle layer, a terrain elevation layer, and a road semantic layer. The corresponding environment map includes an obstacle map, a terrain elevation map, and a road semantic map.

[0196] The obstacle map is mainly used to represent the location and distribution of obstacles in the scene. In some embodiments, the point cloud height of each point cloud can be obtained based on the point cloud data, and then an obstacle map can be constructed based on the height difference between the point cloud heights of any two adjacent point clouds and a preset height difference threshold. Among them, the preset height difference threshold is the maximum height difference between the point cloud heights of two adjacent point clouds at a certain location when there is an obstacle in the current driving environment. Once the height difference between the point cloud heights of two adjacent point clouds is greater than the preset height difference threshold, it means that there is an obstacle here. Therefore, the point cloud height can be obtained based on the point cloud data. Then, binary modeling is performed according to the preset height difference threshold. When the height difference between two adjacent point clouds is higher than the preset height difference threshold, it is an obstacle, otherwise it is a non-obstacle. Then, an obstacle layer map can be constructed based on the information of the point cloud determined to be an obstacle and the point cloud information of non-obstacles.

[0197] A terrain elevation map is a map that shows the ups and downs of the earth's surface. A terrain elevation map uses contour lines, colors, shadows, or other symbols to display the three-dimensional shape of the earth's surface. In some embodiments, based on point cloud data, the point cloud height of each point cloud is obtained, and then a terrain elevation map is constructed based on the point cloud height. The coordinates of the current driving environment in the above embodiments are all provided by the terrain elevation map. In particular, when adjusting the initial path, the vehicle is driving in the current driving environment, so at this time, the vehicle's sensors can be used to collect data on the current driving environment in real time, such as using radar to obtain higher-resolution point cloud data of the environment near the vehicle, such as collecting higher-resolution point cloud data in a semicircular area with a radius of 5 meters in front of the vehicle. Then, the terrain elevation map is updated using the point cloud data collected in real time to obtain a terrain elevation map with higher accuracy. When calculating the dynamic cost subsequently, the calculation can be based on the updated terrain elevation map to improve the accuracy of the dynamic cost.

[0198] The road semantic map not only contains the geometric information of the geographic spatial location, but also contains the semantic information of various objects, roads, traffic rules, etc. in the environment. In some embodiments, image classification can be performed based on the image data to obtain the image semantics corresponding to each area in the image data. Then, a road semantic map is generated based on the image semantics and point cloud data corresponding to each area. Specifically, feature extraction can be performed on the collected image data, for example, using a two-channel convolutional neural network to extract image features. Then, pixel-level image classification is performed to identify dirt roads, grass, sand, shrubs and unknown types in the current driving environment to obtain image semantics. Then, based on the coordinate system conversion relationship between the camera that obtains the image data and the radar that obtains the point cloud data, the image semantics are assigned to the collected point cloud data to obtain a semantic point cloud. Finally, a road semantic map is constructed based on the semantic point cloud.

[0199] In this way, an obstacle map, a terrain elevation map, and a road semantic map of the current driving environment can be constructed, so that the road information of the current driving environment can be accurately obtained based on these three maps, thereby facilitating the improvement of the accuracy of path planning.

[0200] See also Figure 15 The electronic device 100 of the embodiment of the present application includes a processor 20, a memory 30 and a computer program. The computer program is stored in the memory 30 and executed by the processor 20. The computer program includes instructions for executing the path planning method of any of the above embodiments.

[0201] See also Figure 16 The vehicle 200 of the embodiment of the present application includes the electronic device 100 of any of the above embodiments.

[0202] See also Figure 17The embodiment of the present application also provides a computer-readable storage medium 300 on which a computer program 310 is stored. When the computer program 310 is executed by the processor 320, the steps of the path planning method of any of the above-mentioned embodiments are implemented. For the sake of brevity, they are not repeated here.

[0203] See also Figure 18 The embodiment of the present application further provides a computer program product 400, including a computer program 410, wherein the computer program 410 includes a path planning method for executing any of the above embodiments, which will not be described in detail here for the sake of brevity.

[0204] In the description of this specification, the reference terms "certain embodiments", "in an example", "exemplarily", etc. mean that the specific features, structures, materials or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are mutually inconsistent.

[0205] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0206] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A path planning method, characterized in that: Applied to a vehicle, the method comprises: Path planning is performed based on the vehicle's state information and road information to generate a target path; when the vehicle travels along the target path, the vehicle's state information meets preset dynamic cost requirements and / or chassis passability requirements.

2. The path planning method according to claim 1, characterized in that: The performing path planning based on the vehicle state information and road information to generate a target path includes: Plan the initial path based on road information; The initial path is adjusted based on the state information of the vehicle to generate a target path.

3. The path planning method according to claim 2, characterized in that: The adjusting the initial path based on the state information of the vehicle to generate a target path includes: Based on the current path point and the initial path, obtaining multiple candidate paths corresponding to the current path point; When it is determined that the state information of the vehicle meets the dynamic cost requirement and / or the chassis passability requirement, the corresponding candidate path is the target path.

4. The path planning method according to claim 3, characterized in that: The acquiring, based on the current path point and the initial path, a plurality of candidate paths corresponding to the current path point includes: Determine a moving path point corresponding to the current path point in the initial path; Determine a plurality of sampling points corresponding to the moving path point based on a preset sampling rule; According to the current path point and multiple sampling points of the moving path point, multiple candidate paths corresponding to the current path point are determined.

5. The path planning method according to claim 4, characterized in that: The step of determining the sampling points of the moving path points based on a preset sampling rule includes: Determining the number of samples of the sampling point based on the distance between the current path point and the moving path point and a preset offset; A sampling point of the moving path point is determined based on the sampling quantity on a line perpendicular to a connection line between the current path point and the moving path point.

6. The path planning method according to claim 3, characterized in that: When determining that the vehicle meets the dynamic cost requirement and / or the wheel passability requirement, the corresponding candidate path is the target path, including: Determining the dynamic cost of the vehicle when it is located on each candidate path and / or the availability of chassis passability based on the vehicle status information; Determine that the dynamic cost of the vehicle meets the dynamic cost requirement, and / or that the chassis passability of the vehicle meets the chassis passability requirement. A candidate path is the target path.

7. The path planning method according to claim 6, characterized in that: When it is determined that the vehicle meets the dynamic cost requirement and the chassis passability requirement, the corresponding candidate path is the target path, including: Determining a dynamic cost when the vehicle is located on the candidate path based on the current state information of the vehicle and the road information of the candidate path; and / or The chassis passability of the vehicle when located on the to-be-selected path is determined based on the current state information of the vehicle and the road information of the to-be-selected path.

8. The path planning method according to claim 7, characterized in that: Determining a dynamic cost when the vehicle is located on the candidate path according to the current state information of the vehicle and the road information of the candidate path includes: Determining a road surface cost of the candidate path according to the road information of the candidate path; Determining the operation cost of the vehicle according to the current state information of the vehicle and the road information of the selected path; The dynamic cost is determined based on the road surface cost and the operating cost.

9. The path planning method according to claim 8, characterized in that: The road surface cost includes at least one of a target distance between a sampling point of the to-be-selected path and the current path point, and a road surface semantic cost.

10. The path planning method according to claim 9, characterized in that: In the case where the road surface cost includes a road surface semantic cost, the road surface cost of the candidate path is determined according to the road information of the candidate path, including: Determining the proportions of different road surface types in the to-be-selected path based on the road information; The road semantic cost is determined based on the proportion of different road types in the candidate path and the weight coefficient corresponding to each road type.

11. The path planning method according to claim 10, characterized in that: The road information includes a road semantic map corresponding to the current driving environment, and determining the proportion of different road types in the candidate path based on the road information includes: The proportions of different road surface types in the candidate path are determined based on the road surface semantic map.

12. The path planning method according to claim 8, characterized in that: The operation cost includes at least one of a dynamic stability cost and a static stability cost of the vehicle.

13. The path planning method according to claim 12, wherein: In a case where the operating cost includes a dynamic stability cost of the vehicle, the current state information includes a current speed of the vehicle, and the road information includes a curvature radius corresponding to each node on a to-be-selected path, determining the operating cost of the vehicle based on the current state information of the vehicle and the road information of the to-be-selected path includes: Determining the lateral acceleration corresponding to each node according to the current speed and the curvature radius corresponding to each node, wherein the plurality of nodes divide the path to be selected; The dynamic stability cost is determined according to the lateral acceleration corresponding to each of the nodes.

14. The path planning method according to claim 12, characterized in that: In a case where the operation cost includes a static stability cost of the vehicle, the road information includes a pitch angle and a roll angle of the current path point, and a pitch angle and a roll angle of a sampling point corresponding to the to-be-selected path, and determining the operation cost of the vehicle based on the current state information of the vehicle and the road information of the to-be-selected path includes: The static stability cost is determined according to the pitch angle and the roll angle of the current path point and the pitch angle and the roll angle of the sampling point corresponding to the selected path.

15. The path planning method according to claim 8, characterized in that: The determining the dynamic cost based on the road surface cost and the operation cost includes: The dynamic cost is determined based on the road surface cost, the weight coefficient corresponding to the road surface cost, the operation cost, and the weight coefficient corresponding to the operation cost.

16. The path planning method according to claim 7, characterized in that: The determining, based on the current state information of the vehicle and the road information of the to-be-selected path, whether the chassis of the vehicle is capable of carrying traffic when the vehicle is located on the to-be-selected path includes: Substituting the current state information and the road information of the candidate path into a preset passability equation to obtain a passability parameter, wherein the passability parameter includes a height difference between the ground of the candidate path and the chassis of the vehicle; When the passability parameter satisfies a preset passability condition, it is determined that the vehicle has chassis passability when located on the to-be-selected path.

17. The path planning method according to claim 16, characterized in that: The determining, based on the current state information of the vehicle and the road information of the candidate path, whether the chassis of the vehicle is capable of passing the vehicle when the vehicle is located on the candidate path, further includes: When the passability parameter does not satisfy the preset passability condition, the vehicle does not have chassis passability when located on the to-be-selected path.

18. The path planning method according to claim 17, characterized in that: The preset passability condition includes the passability parameter being less than a preset value.

19. The path planning method according to claim 18, characterized in that: The preset value is zero.

20. The path planning method according to claim 16, wherein: Substituting the current state information and the road information of the selected path into a preset passability equation to obtain passability parameters, including: Substituting the current state information and the coordinates of each node of the to-be-selected path in the current driving environment into the preset passability equation to obtain the passability parameters of each node; When the passability parameter satisfies a preset passability condition, determining that the vehicle has chassis passability when located on the to-be-selected path includes: When the passability parameters corresponding to all nodes of the to-be-selected path satisfy preset passability conditions, it is determined that the vehicle has chassis passability.

21. The path planning method according to claim 16, wherein: The current state information includes at least the vehicle's four-wheel suspension height, the vehicle's location, and one of the vehicle's body pitch angle and roll angle; and / or the road information of the selected path includes the coordinates of the selected path in the current driving environment.

22. The path planning method according to any one of claims 1 to 21, characterized in that: The dynamic cost requirement includes minimizing the dynamic cost, and the chassis passability requirement includes that the vehicle has chassis passability.

23. The path planning method according to any one of claims 1 to 21, characterized in that: The initial path planning based on the road information includes: Determining, based on the current planned path point and the road information, a planned partial path corresponding to the current planned path point, wherein a static cost corresponding to the planned partial path meets a static cost requirement, and the current planned path point at least includes a starting point of the current driving path; The end point of the planned partial path is used as the next current planned path point, and the step of determining the planned partial path corresponding to the current planned path point based on the current planned path point and the road information is entered again until the planned partial path includes the end point of the trip in the current driving environment; The initial path is determined according to all planned local paths of the current driving environment.

24. The path planning method according to claim 23, characterized in that: The determining, based on the current planned path point and the road information, a planned partial path corresponding to the current planned path point includes: Determining a plurality of candidate path points corresponding to the currently planned path point based on the currently planned path point and the road information; Determining a static cost of each candidate path point according to the road information corresponding to each candidate path point; Based on the candidate path points whose static costs meet the static cost requirement, a selected path point of the current planned path point is determined to determine the planned local path of the current planned path point.

25. The path planning method according to claim 24, characterized in that: The road information includes an obstacle map and a terrain elevation map of the current driving environment; The determining, based on the current planned path point and the road information, a plurality of candidate path points corresponding to the current planned path point includes: Obtaining a target moving distance based on a current position of the currently planned path point and the obstacle map; Determine multiple initial path points corresponding to the current planned path point based on the target moving distance, the current planned path point and a preset angle; Determine a plurality of candidate path points corresponding to the currently planned path point based on the target moving distance, the initial path point, and the terrain elevation map.

26. The path planning method according to claim 25, characterized in that: The acquiring of the target moving distance based on the current position of the currently planned path point and the obstacle map includes: Based on the obstacle map, obtaining the closest distance between the current position of the currently planned path point and the obstacle, and using the closest distance as the initial moving distance; A target moving distance is determined based on the initial moving distance and a preset distance range.

27. The path planning method according to claim 25, characterized in that: The determining, based on the target moving distance, the initial path point, and the terrain elevation map, a plurality of candidate path points corresponding to the currently planned path point includes: Based on the target initial path point and the terrain elevation map, obtaining an intermediate path point corresponding to the target initial path point on the terrain elevation map, wherein the target initial path point is any of the initial path points; Based on the target moving distance, the height of the intermediate path point in the terrain elevation map and the height of the currently planned path point in the terrain elevation map, the candidate path point corresponding to the intermediate path point is determined to determine multiple candidate path points corresponding to the currently planned path point.

28. The path planning method according to claim 27, characterized in that: The determining, based on the target moving distance, the height of the intermediate path point on the terrain elevation map, and the height of the currently planned path point on the terrain elevation map, the candidate path point corresponding to the intermediate path point includes: Obtaining a unit climbing height between the intermediate path point and the current planned path point according to the height of the intermediate path point on the terrain elevation map and the height of the current planned path point on the terrain elevation map; Determining a dynamic moving distance based on the unit's climbing height and the target's moving distance; The candidate path point corresponding to the intermediate path point is determined based on the current planned path point, the dynamic moving distance and the direction of the intermediate path point, where the direction of the intermediate path point is the direction of the current planned path point toward the intermediate path point.

29. The path planning method according to claim 24, characterized in that: The determining, based on the road information corresponding to each candidate path point, a static cost of each candidate path point includes: Determining a movement cost of a target candidate path point based on the road information corresponding to the current planned path point and the target candidate path point, where the target candidate path point is any of the candidate path points; Determining a heuristic cost of the target candidate pathpoint based on the road information corresponding to the target candidate pathpoint and the road information corresponding to the end point of the trip in the current driving environment; The static cost of the target candidate path point is determined according to the movement cost and the heuristic cost.

30. The path planning method according to claim 29, characterized in that: The movement cost of the target candidate path point includes at least one of the movement cost of the current planned path point, the Euclidean distance between the current planned path point and the target candidate path point, the static stability cost and the road semantic cost.

31. The path planning method according to claim 30, characterized in that: In a case where the movement cost includes the static stability cost, determining the movement cost of the target candidate path point according to the road information corresponding to the current planned path point and the target candidate path point includes: The static stability cost of the target candidate path point is determined based on the pitch angle of the current planned path point and the pitch angle of the target candidate path point, and / or the roll angle of the current planned path point and the roll angle of the target candidate path point.

32. The path planning method according to claim 30, characterized in that: In a case where the movement cost includes the road semantic cost, the road information includes a road semantic map corresponding to the current driving environment, and determining the movement cost of the target candidate path point based on the road information corresponding to the current planned path point and the target candidate path point includes: Determining, based on the road semantic map, the proportions of different road surface types in the path between the target candidate path point and the currently planned path point; The road semantic cost is determined based on the proportion of different road surface types in the path between the target candidate path point and the currently planned path point and the weight coefficient corresponding to each of the road surface types.

33. The path planning method according to claim 29, characterized in that: The heuristic cost includes the Euclidean distance between the end point of the current driving environment and the target candidate path point.

34. The path planning method according to claim 23, wherein: The static cost requirement includes minimizing the static cost.

35. The path planning method according to claim 1, characterized in that: The method further comprises: An environment map corresponding to the current driving environment is determined based on preset image data and preset point cloud data of the current driving environment, where the environment map includes the road information.

36. The path planning method according to claim 35, characterized in that: The environment map includes at least one or more of an obstacle map, a terrain elevation map, and a road semantic map.

37. The path planning method according to claim 36, characterized in that: In a case where the environmental map includes a road semantic map, determining the environmental map corresponding to the current driving environment based on the image data and the point cloud data of the current driving environment includes: performing image classification based on the image data to obtain image semantics corresponding to each region in the image data; A road semantic map is generated according to the image semantics and the point cloud data corresponding to each area.

38. The path planning method according to claim 36, characterized in that: In a case where the environmental map includes a terrain elevation map, determining the environmental map corresponding to the current driving environment based on the image data and the point cloud data of the current driving environment includes: Based on the point cloud data, obtaining the point cloud height of each point cloud; The terrain elevation map is constructed based on the point cloud height.

39. The path planning method according to claim 36, characterized in that: In a case where the environment map includes an obstacle map, determining the environment map corresponding to the current driving environment based on the image data and point cloud data of the current driving environment includes: Based on the point cloud data, obtaining the point cloud height of each point cloud; The obstacle map is constructed based on the height difference between the point cloud heights of any two adjacent point clouds and a preset height difference threshold.

40. An electronic device, characterized in that: include: Processor, memory; and A computer program, wherein the computer program is stored in the memory and executed by the processor, and the computer program includes instructions for executing the path planning method according to any one of claims 1 to 39.

41. A vehicle, characterized in that include: The electronic device of claim 40.

42. A non-volatile computer-readable storage medium containing a computer program, characterized in that When the computer program is executed by a processor, the processor executes the path planning method according to any one of claims 1 to 39.

43. A computer program product, characterized in that The computer program comprises a computer program for executing the path planning method according to any one of claims 1 to 39.

Images