Automatic driving vehicle speed planning method and device, electronic equipment and medium

Abstract

The present disclosure relates to an automatic driving vehicle speed planning method, device, electronic equipment and medium. The method comprises: determining a drivable area of a vehicle to be speed planned in a space-time relationship graph, the space-time relationship graph being associated with a planned driving track of the vehicle; determining a plurality of nodes in the space-time relationship graph, each node of the plurality of nodes indicating a corresponding spatial position of the vehicle at a specific time in the planned driving track; determining a node at a next time in the drivable area based on a node corresponding to the vehicle at a current time, to obtain a space-time curve corresponding to the planned driving track in the space-time relationship graph; and obtaining a target planning speed of the vehicle at different times based on the space-time curve. In this way, automatic driving speed planning can be efficiently realized with low computational complexity, hardware cost is reduced, and driving safety and driving experience are guaranteed.

Description

Technical Field

[0001] This disclosure generally relates to the field of autonomous driving technology, and particularly to methods, apparatus, electronic devices, and computer-readable storage media for speed planning of autonomous vehicles. Background Technology

[0002] With the development of autonomous driving technology, more and more functions related to autonomous driving technology are being applied to mass-produced vehicles. The development of this technology is also an important means to alleviate traffic congestion and enhance highway safety in the future.

[0003] Local path planning is a crucial function of autonomous driving technology. It plans the local path and speed trajectory of an autonomous vehicle based on its global path, position, and predicted trajectories of surrounding obstacles. Typically, local path planning involves two steps: first, a local path to the target lane is planned based on the vehicle's global path, position, and predicted trajectories of surrounding obstacles; then, based on the planned local path and predicted trajectories of surrounding obstacles, the vehicle's speed at each waypoint is planned, under constraints such as road speed limits and vehicle comfort.

[0004] Autonomous vehicle speed planning algorithms typically determine the obstacles involved in the interaction based on the planned local path, discretize it into a space-time relationship graph, project the space and time occupied by the obstacles onto the space-time relationship graph to determine the drivable area of ​​the autonomous vehicle, and then recursively deduce all drivable nodes of the autonomous vehicle under constraints such as road speed limits and vehicle comfort from the time and space location of the autonomous vehicle. Then, find the optimal node from the drivable nodes at the last moment, and backtrack from the node to find the optimal space-time curve to determine the driving speed of the autonomous vehicle to each path point.

[0005] This method requires a lot of computation during the backward recursion from the current position and the forward backtracking from the final position. It has low speed planning efficiency, which increases the hardware cost of autonomous vehicles, affects the comfort of autonomous driving, and in severe cases, it can also affect the safety of autonomous driving. Summary of the Invention

[0006] According to an example embodiment of this disclosure, an autonomous vehicle speed planning scheme is provided to at least partially address the problems existing in the prior art.

[0007] In a first aspect of this disclosure, a speed planning method for an autonomous vehicle is provided. The method includes: determining a drivable area of ​​the vehicle to be speed planned in a space-time graph, the space-time graph being associated with the vehicle's planned driving trajectory; determining multiple nodes in the space-time graph, each node indicating the corresponding spatial position of the vehicle in the planned driving trajectory at a specific time; determining nodes within the drivable area at the next time based on the node corresponding to the vehicle at the current time, to obtain a space-time curve corresponding to the planned driving trajectory in the space-time graph; and obtaining the target planned speed of the vehicle at different times based on the space-time curve.

[0008] In some embodiments, determining the drivable area of ​​the vehicle to be speed-planned in the space-time relationship diagram may include: establishing a space-time relationship diagram based on the planned driving trajectory of the vehicle; projecting the space and time occupied by one or more obstacles onto the space-time relationship diagram to obtain the non-drivable area of ​​the vehicle, wherein the predicted driving paths of one or more obstacles intersect the planned driving trajectory of the vehicle at least partially; and determining the area outside the non-drivable area in the space-time relationship diagram as the drivable area.

[0009] In some embodiments, determining multiple nodes in a spatial-temporal relationship diagram may include: discretizing the spatial-temporal relationship diagram into multiple squares at preset intervals; and determining the center point of each square in the multiple squares as multiple nodes.

[0010] In some embodiments, determining the nodes within the drivable area at the next moment based on the node corresponding to the vehicle at the current moment, so as to obtain a space-time curve corresponding to the planned driving trajectory in the space-time relationship graph, may include: selecting the node indicating the spatial position of the vehicle at the current moment as the parent node and recursively pushing to the next moment; calculating the cost of the drivable child node at the next moment according to the cost function; selecting the child node with the minimum cost as the updated parent node and continuing to recursively push to the subsequent moments of the next moment; and in response to the successful determination of the recursion between the nodes at each moment, connecting a series of child nodes to determine the space-time curve.

[0011] In some embodiments, selecting the child node with the lowest cost as the updated parent node and continuing to recursively push forward to subsequent time steps may include: determining the vehicle's comfort constraints and road speed limits, wherein the comfort constraints are associated with one or more of the vehicle's longitudinal velocity, acceleration, jerk, and lateral acceleration; in response to determining that each child node recursively pushed forward from the updated parent node does not satisfy either the comfort constraints or the road speed limits, reselecting the child node with the second lowest cost as the updated parent node and recursively pushing forward to subsequent time steps; and in response to determining that each child node recursively pushed forward from the second lowest cost child node as the updated parent node does not satisfy either the comfort constraints or the road speed limits, reselecting the child node with the third lowest cost as the updated parent node and recursively pushing forward to subsequent time steps.

[0012] In some embodiments, selecting the child node with the lowest cost as the updated parent node and continuing to recursively push forward to the next time step may include: traversing the current time step and recursively pushing forward from the parent node to each child node until neither the comfort constraint nor the road speed limit is satisfied, backtracking to the previous time step; and selecting the child node with the second lowest cost at the previous time step and recursively pushing forward to the next time step until the recursively pushed node satisfies both the comfort constraint and the road speed limit.

[0013] In some embodiments, obtaining the target planned speed of a vehicle at different times based on a space-time curve may include: determining the slope of each node on the space-time curve as the target planned speed; optionally or additionally, performing secondary smoothing on the space-time curve to calculate the speed of each node on the space-time curve; and binding the speed of each node with the corresponding trajectory point on the planned driving trajectory to obtain the target planned speed.

[0014] In a second aspect of this disclosure, an autonomous vehicle speed planning apparatus is provided. The apparatus includes: a drivable area determination module configured to determine a drivable area of ​​the vehicle to be speed planned in a space-time relationship graph, the space-time relationship graph being associated with the planned driving trajectory of the vehicle; a node determination module configured to determine multiple nodes in the space-time relationship graph, each of the multiple nodes indicating a corresponding spatial position of the vehicle in the planned driving trajectory at a specific time; a space-time curve determination module configured to determine nodes within the drivable area at the next time moment based on the nodes corresponding to the vehicle at the current time moment, so as to obtain a space-time curve corresponding to the planned driving trajectory in the space-time relationship graph; and a target planned speed acquisition module configured to acquire the target planned speed of the vehicle at different times based on the space-time curve.

[0015] In some embodiments, the drivable area determination module may also be configured to: establish a spatial-temporal relationship diagram based on the planned driving trajectory of the vehicle; project the space and time occupied by one or more obstacles onto the spatial-temporal relationship diagram to obtain the non-drivable area of ​​the vehicle, wherein the predicted driving paths of one or more obstacles intersect the planned driving trajectory of the vehicle at least partially; and determine the area outside the non-drivable area in the spatial-temporal relationship diagram as the drivable area.

[0016] In some embodiments, the node determination module may also be configured to discretize the spatial-temporal relationship diagram into multiple squares at preset intervals; and to determine the center point of each square in the multiple squares as multiple nodes.

[0017] In some embodiments, the space-time curve determination module can also be configured to select a node indicating the spatial position of the vehicle at the current moment as the parent node and recursively push to the next moment; calculate the cost of the drivable child node at the next moment according to the cost function; select the child node with the minimum cost as the updated parent node and continue to recursively push to the subsequent moments of the next moment; and in response to the successful determination of the recursion between the nodes at each moment, determine a series of child nodes as a space-time curve by connecting them.

[0018] In some embodiments, the space-time curve determination module can also be configured to determine the vehicle's comfort constraints and road speed limits, the comfort constraints being associated with one or more of the vehicle's longitudinal velocity, acceleration, jerk, and lateral acceleration; in response to determining that each child node recursively from the updated parent node does not satisfy either the comfort constraints or the road speed limits, the second least costly child node is reselected as the updated parent node for recursion to subsequent time points; and in response to determining that each child node recursively from the second least costly child node as the updated parent node for recursion to subsequent time points does not satisfy either the comfort constraints or the road speed limits, the third least costly child node is reselected as the updated parent node for recursion to subsequent time points.

[0019] In some embodiments, the space-time curve determination module can also be configured to iterate through the current moment from the parent node to each child node, and if none of them satisfy either the comfort constraint or the road speed limit, backtrack to the previous moment; and select the child node with the second smallest cost at the previous moment to recursively push to subsequent moments until the recursed node satisfies both the comfort constraint and the road speed limit.

[0020] In some embodiments, the target planned speed acquisition module may also be configured to determine the slope of each node on the space-time curve as the target planned speed, and optionally or additionally, to perform secondary smoothing on the space-time curve to calculate the speed of each node on the space-time curve, and to bind the speed of each node to the corresponding trajectory point on the planned driving trajectory to obtain the target planned speed.

[0021] In a third aspect of this disclosure, an electronic device is provided. The device includes: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the method according to a first aspect of this disclosure.

[0022] In a fourth aspect of this disclosure, a computer-readable storage medium is provided. The medium stores a computer program that, when executed by a processor, implements the method according to a first aspect of this disclosure.

[0023] In a fifth aspect of this disclosure, a computer program product is provided. The product includes a computer program / instructions that, when executed by a processor, implement the method according to a first aspect of this disclosure.

[0024] It should be understood that the description in the Summary of the Invention is not intended to limit the key or essential features of the embodiments of this disclosure, nor is it intended to restrict the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0025] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements. The drawings are provided for a better understanding of the present invention and are not intended to limit the scope of this disclosure, wherein:

[0026] Figure 1 A schematic diagram of an example environment in which several embodiments of the present disclosure can be implemented is shown;

[0027] Figure 2 A schematic flowchart of an autonomous vehicle speed planning method according to some embodiments of the present disclosure is shown;

[0028] Figure 3 The diagram shows a spatial-temporal relationship diagram and a schematic diagram of the spatial-temporal relationship occupied by obstacles, according to some embodiments of the present disclosure.

[0029] Figure 4 This diagram illustrates a recursive approach from parent node to child node according to some embodiments of the present disclosure;

[0030] Figure 5 This diagram illustrates the selection of the second least-cost sub-node when recursion is not possible according to some embodiments of this disclosure;

[0031] Figure 6 This diagram illustrates the selection of the third least-cost sub-node when recursion is not possible according to some embodiments of this disclosure;

[0032] Figure 7 This diagram illustrates the selection of a node backtracking to the previous time step when none of the child nodes can be recursively deduced according to some embodiments of the present disclosure.

[0033] Figure 8 An overall flowchart of a speed planning method according to some embodiments of the present disclosure is shown;

[0034] Figure 9 A schematic block diagram of an autonomous vehicle speed planning device according to some embodiments of the present disclosure is shown; and

[0035] Figure 10 A block diagram of a computing device capable of implementing several embodiments of the present disclosure is shown. Detailed Implementation

[0036] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0037] In the description of embodiments of this disclosure, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0038] During the operation of an autonomous vehicle, a driving trajectory to the target lane is planned based on its global path, position, and predicted trajectory information of surrounding obstacles. There will be an intersection area between the predicted trajectory of the autonomous vehicle and the predicted trajectory of surrounding obstacles (such as obstacle vehicles driving on the road). This intersection area is the obstacle area of ​​the autonomous vehicle.

[0039] The autonomous vehicle speed planning method can identify the obstacles involved in the interaction based on the planned local path, discretize them into a space-time relationship graph, project the space and time occupied by the obstacles onto the space-time relationship graph to determine the drivable area of ​​the autonomous vehicle, and then recursively deduce all drivable nodes of the autonomous vehicle under constraints such as road speed limits and vehicle comfort from the time and space location of the autonomous vehicle. Then, find the optimal node from the drivable nodes at the last moment, and backtrack from the node to find the optimal space-time curve to determine the driving speed of the autonomous vehicle to each path point.

[0040] As mentioned earlier, this speed planning method requires first traversing all drivable nodes from the initial position to the final position in the space-time graph to find the optimal node. Then, it backtracks from the final position to find the optimal space-time curve. Furthermore, it needs to determine the speed of the autonomous vehicle at each waypoint. Therefore, the computational load is huge, and the hardware requirements are high. At the same time, due to the low planning efficiency, it affects the comfort of the autonomous vehicle and may also lead to driving safety issues in scenarios where rapid decision-making is required at high speeds.

[0041] To address the above problems, the solutions of various embodiments of this disclosure establish a spatial-temporal relationship graph based on the planned driving trajectory of the autonomous vehicle. The space and time occupied by obstacles are projected onto the spatial-temporal relationship graph to determine the drivable area of ​​the autonomous vehicle. The spatial-temporal relationship graph is discretized to distinguish between drivable and non-drivable areas. According to the various embodiments of this disclosure, the drivable child nodes at the next moment are obtained by recursively calculating the cost of each drivable child node, selecting the child node with the lowest cost as the new parent node, and repeating the above steps. If the recursion is successful at each moment in the spatial-temporal relationship graph, the resulting series of child nodes connected together form a spatial-temporal curve, and the slope of the curve at each node represents the speed value of that node, indicating successful speed planning. If, after recursing to a certain node, none of its child nodes meet constraints such as road speed limits and vehicle comfort, and further recursion is impossible, then speed planning fails. When recursion fails at this moment, the child node with the lowest cost is selected as the new parent node for recursion. If the second smallest child node of the cost still cannot be recursively pushed forward, then the third smallest child node of the cost is selected as the parent node to push forward the cost, and the above steps are repeated. If all child nodes at this moment fail to be pushed forward and cannot be pushed forward, then backtrack to the previous moment, select the second smallest child node of the cost at this moment again to push forward, and repeat the above steps until the push is successful and the speed planning is successful.

[0042] In this way, the speed planning method only needs to consider the cost of each node at the current moment when selecting a specific node, without considering the costs of drivable and indestructible nodes at other moments, and without calculating the cost of all nodes in the entire space-time graph for speed planning. Furthermore, when node selection fails at the current moment, tracing back only requires considering the nodes at the previous moment, thus significantly reducing the computational load of autonomous driving path speed planning and lowering hardware requirements. Moreover, this method significantly improves speed planning efficiency, thereby ensuring driving safety and a better driving experience.

[0043] The following will combine Figures 1 to 10 Exemplary embodiments of this disclosure are described below.

[0044] Figure 1 A schematic diagram of an example environment 100 in which several embodiments of the present disclosure can be implemented is shown.

[0045] like Figure 1 As shown, in the example of environment 100, vehicle 101 is driving on the road. It should be understood that... Figure 1 The environment 100 shown is only one example environment in which the vehicle 101 may be driving. In addition to driving on outdoor roads, the vehicle 101 may also drive in various environments such as tunnels, outdoor parking lots, inside buildings (e.g., indoor parking lots), residential areas, parks, etc.

[0046] exist Figure 1 In the example, vehicle 101 can be any type of vehicle capable of carrying people and / or goods and moving via a power system such as an engine, including but not limited to cars, trucks, buses, electric vehicles, motorcycles, RVs, trains, etc. In some embodiments, vehicle 101 (also referred to as vehicle 101) in environment 100 can be a vehicle with a certain degree of autonomous driving capability; such a vehicle is also referred to as a driverless vehicle or an autonomous vehicle. In some embodiments, vehicle 101 can also be a vehicle with semi-autonomous driving capability.

[0047] Bicycle 101 in Ru Figure 1 When driving in the lane shown, the vehicle 101 has an initial planned path. Normally, the vehicle 101 drives normally along the initial planned path. However, when an obstacle exists in the initial planned path, the vehicle 101 typically modifies the initial planned path using the computing device 110 to obtain an optimized planned path. In one embodiment, the obstacle may be, for example, an obstacle vehicle 103 that is also driving in the same lane as the vehicle 101. The type of obstacle vehicle 103 can be any type of vehicle 101 as described above, and the obstacle vehicle 103 can also be a vehicle without autonomous driving capabilities; this disclosure does not limit this.

[0048] Continue to refer to Figure 1 When the obstacle vehicle 103 is on the initial planned path of the vehicle 101, or when the predicted path of the obstacle vehicle intersects with the initial planned path of the vehicle 101, the obstacle vehicle 103 will hinder the normal driving of the vehicle 101. In this case, the initial planned path is modified by the calculation device 110 into an optimized planned path. When the vehicle 101 is driving along the optimized planned path, it also needs to consider how to plan the speed of each trajectory point on the optimized path, so as to ensure that the road speed limit requirements are met while also taking into account driving comfort and safety. This can also be achieved by the calculation device 110.

[0049] like Figure 1 As shown, computing device 110 can be communicatively coupled to vehicle 101. Although shown as a separate entity, computing device 110 can be embedded within vehicle 101. Computing device 110 can also be an entity external to vehicle 101 and can communicate with vehicle 101 via a wireless network. Computing device 110 can be any device with computing capabilities.

[0050] As a non-limiting example, computing device 110 can be any type of fixed computing device, mobile computing device, or portable computing device, including but not limited to desktop computers, laptop computers, notebook computers, netbook computers, tablet computers, multimedia computers, mobile phones, etc.; all or some of the components of computing device 110 can be distributed in the cloud. Computing device 110 includes at least a processor, memory, and other components typically found in general-purpose computers to perform computing, storage, communication, and control functions.

[0051] In some embodiments, the computing device 110 may include a system for speed planning of an autonomous vehicle. The system may include a perception prediction module, a map service module, a positioning module, a decision planning module, and a control execution module. The perception prediction module includes, but is not limited to, cameras and radar, for acquiring information such as the position and speed of obstacles (e.g., obstacle vehicle 103) around the vehicle 101 and lane line information. Simultaneously, the perception prediction module can also predict the predicted paths of obstacles, such as the predicted path of obstacle vehicle 103, so that the vehicle 101 can determine whether its initial planned path intersects with the predicted paths of obstacles. The map service module is equipped with a high-precision map and can provide a global path for the autonomous vehicle 101. The positioning module can obtain high-precision global positioning information of the vehicle using a global positioning system. The decision planning module can, based on the global path of the high-precision map, comprehensively analyze and plan a local path for the vehicle by combining the perception and positioning information. The control execution module can control the vehicle to travel according to the planned path. It should be understood that the methods according to the various embodiments of this disclosure can be executed by the decision planning module, and the various modules mentioned above can be at least partially included in the computing device 110.

[0052] Therefore, the computing device 110 can plan the local path trajectory and local speed trajectory of the autonomous vehicle based on information such as the global path, position, and predicted trajectories of surrounding obstacles. The speed of the vehicle 101 at each path point on the local path trajectory can be planned according to various embodiments of this disclosure.

[0053] Figure 2 A schematic flowchart of an autonomous vehicle speed planning method 200 according to some embodiments of the present disclosure is shown. Method 200 may, for example, be derived from... Figure 1 The computing device 110 shown is implemented.

[0054] In box 201, the drivable area of ​​the vehicle to be speed-planned is determined in the space-time relationship diagram, which is associated with the vehicle's planned driving trajectory. The vehicle is, for example, like... Figure 1 The following description uses the autonomous vehicle 101 as an example. The spatial-temporal relationship diagram can be established based on the planned driving trajectory of the autonomous vehicle, and it can reflect the specific spatial-temporal position of the autonomous vehicle 101. In one embodiment, the spatial-temporal relationship diagram can be... Figure 1 The implementation may be carried out in the computing device 110 shown, or it may be carried out in other devices independent of the computing device 110, or partially in the computing device 110 and partially in other devices. This disclosure does not limit this.

[0055] A spatial-temporal relationship diagram can be, for example, as follows: Figures 3 to 7 The coordinate system shown is implemented. See also Figures 3 to 7The horizontal axis of the coordinate system represents the various times T on the planned driving trajectory of the vehicle 101, while the vertical axis represents the spatial position S that the vehicle 101 can achieve at a specific time. For example, the vehicle 101 can determine the driving speed at each trajectory point of the planned driving trajectory by selecting the spatial position S at a specific time through the computing device 110, which will be described in detail below.

[0056] In the time-space relationship diagram, when there are no obstacles (such as obstacle vehicle 103) around vehicle 101, the vehicle can drive to any area in the time-space relationship diagram, which is the drivable area of ​​vehicle 101. However, when there are obstacles (such as obstacle vehicle 103), part of the area in the time-space relationship diagram will be occupied by the obstacles, such as... Figure 2 As shown, this area is considered a non-driving or impassable area, also known as an obstacle area. Figure 2 The obstacle area in the text can be, for example, Figure 1 The projection of the space and time occupied by the obstacle vehicle 103 on the space-time relationship diagram.

[0057] Accordingly, in one embodiment, the drivable area of ​​the vehicle to be speed-planned can be determined in the space-time relationship diagram as follows: based on the planned driving trajectory of the vehicle, a space-time relationship diagram is established, and then the space and time occupied by obstacles are projected onto the space-time relationship diagram to obtain the non-drivable area of ​​the vehicle, wherein the predicted driving paths of one or more obstacles intersect at least partially with the planned driving trajectory of the vehicle, and the area outside the non-drivable area in the space-time relationship diagram is determined as the drivable area.

[0058] In one embodiment, although as Figure 2 The diagram shows only one obstacle, but the number of obstacles can be multiple, such as two, three, five, or any other suitable number; this disclosure does not limit this. It should be understood that the strategy for speed planning of one obstacle and the strategy for speed planning of multiple obstacles can be the same.

[0059] In box 203, multiple nodes are defined in the space-time relationship diagram, each node indicating the corresponding spatial position of the vehicle at a specific moment in the planned driving trajectory. In one embodiment, such as Figures 3 to 7 As shown, the space-time relationship diagram can be discretized into multiple squares at preset intervals, and then the center point of each square in the multiple squares can be determined as multiple nodes. The preset interval can be limited according to road speed limits and vehicle comfort constraints such as vehicle longitudinal speed, acceleration, jerk, and lateral acceleration, or it can be set according to any other suitable method.

[0060] In box 205, the nodes in the drivable area at the next moment are determined based on the nodes corresponding to the vehicle at the current moment, so as to obtain the space-time curve corresponding to the planned driving trajectory in the space-time relationship diagram.

[0061] Figure 4 This illustration shows a recursive diagram from parent node to child node according to some embodiments of the present disclosure. In one embodiment, such as... Figure 4 As shown, the node indicating the spatial position of vehicle 101 at the current moment can be selected as the parent node for recursion to the next moment. For example, vehicle 101 can be at the origin of the space-time relationship graph, corresponding to time t, and then the node selection can be performed at the next time t+1. For example, the cost of the drivable child nodes at the next moment can be calculated according to the cost function, and the child node with the lowest cost can be selected as the updated parent node to continue recursively to the subsequent moments of the next moment. Figure 4 As shown, vehicle 101 selects the third node with the lowest cost at the next time step as the next parent node and recursively moves forward. When the next parent node is recursively moved to the time step after the next time step, it also performs the calculation of the lowest cost and selects the child node as the next updated parent node. Repeating the above steps, if the recursion is successful at each time step in the space-time relationship diagram, the resulting series of child node connections will form the space-time curve.

[0062] In one embodiment, the determination of whether the backward recursion is successful depends on factors such as the vehicle's comfort constraints and road speed limits. Comfort constraints may include parameters such as vehicle longitudinal speed, acceleration, jerk, and lateral acceleration, while the road speed limit represents the maximum permissible speed on the current road segment. Figure 4 In the illustrated embodiment, if after recursively reaching a certain node, none of its child nodes meet constraints such as road speed limits and vehicle comfort, then recursion cannot continue and speed planning fails.

[0063] In this way, when speed planning fails due to constraints such as vehicle comfort and road speed limits, it can be done as follows: Figure 5 The example shown reselects child nodes. Figure 5 A schematic diagram illustrating the selection of the second least-cost child node when recursion is not possible according to some embodiments of the present disclosure is shown.

[0064] In one embodiment, in response to determining that each child node recursively from the updated parent node does not satisfy either the comfort constraint or the road speed limit, the child node with the second smallest cost can be reselected as the updated parent node for subsequent time steps.

[0065] Reference Figure 5In this embodiment, if the selection of the child node with the lowest cost fails at this moment, the second lowest cost child node is selected as the new parent node and the process continues until either the comfort constraint or the road speed limit is no longer satisfied.

[0066] Figure 6 A schematic diagram illustrating the selection of the third least-cost child node when recursion is not possible according to some embodiments of the present disclosure is shown.

[0067] In one embodiment, such as Figure 6 As shown, in response to the situation where the child nodes determined to be the second least costly child node and subsequently recursively updated to parent nodes do not satisfy either the comfort constraint or the road speed limit, the third least costly child node can be reselected as the updated parent node for recursion to subsequent times. In this embodiment, referring to... Figure 6 If the second smallest child node of the cost still cannot be pushed forward, then the third smallest child node of the cost is selected as the parent node and pushed forward until a space-time curve that can avoid obstacles is obtained.

[0068] In one embodiment, when iterating through all child nodes at the current moment fails to successfully proceed, i.e., when iterating from the parent node to each child node at the current moment fails to satisfy either the comfort constraint or the road speed limit, the process backtracks to the previous moment and selects the child node with the second smallest cost at the previous moment to recursively proceed to subsequent moments until the recursed node satisfies both the comfort constraint and the road speed limit. It should be understood that the above operation can be repeated until speed planning is successful.

[0069] Figure 7 This diagram illustrates the selection of a node after backtracking to the previous time step when none of the child nodes can be recursively deduced according to some embodiments of this disclosure. Figure 7 As shown, if it fails to traverse all child nodes at the current time step, backtrack to the previous time step and select the child node with the second smallest cost at the current time step to proceed. If it still fails, select the child node with the third smallest cost at the current time step to proceed.

[0070] In the above embodiments, only the drivable nodes at the current moment can be considered in each operation step, without considering all nodes in the space-time relationship graph, which significantly reduces the amount of computation, improves speed planning efficiency, and enhances autonomous driving safety.

[0071] Subsequently, in box 207, the target planned speed of the vehicle at different times is obtained based on the space-time curve. In one embodiment, the slope of each node on the space-time curve is determined as the target planned speed. Optionally or additionally, the space-time curve can be smoothed twice to calculate the speed of each node on the space-time curve, and the speed of each node is bound to the corresponding trajectory point on the planned driving trajectory to finally obtain the target planned speed.

[0072] In one embodiment, a Bessel interpolation smoothing algorithm can be used for secondary smoothing to recalculate the speed of each child node, ensuring comfort during vehicle operation. Subsequently, the calculated speeds of the child nodes are bound to local path trajectory points, thus planning the driving speed of the autonomous vehicle to each path point, completing the local path speed planning. It should be understood that any other suitable algorithm can be used for secondary smoothing, and this disclosure does not impose any limitations on this.

[0073] Figure 8 A general flowchart of a speed planning method according to some embodiments of the present disclosure is shown. It should be understood that... Figure 8 The flowchart shown illustrates a complete autonomous driving speed planning scheme, in which one or more operations can be omitted or replaced under certain circumstances.

[0074] like Figure 8 As shown, velocity planning begins at t=0. Then, in the space-time graph, the time and space origin are used as the first parent node, and the process iterates backwards for dt time units. Generally, the magnitude of dt is positively correlated with the distance between nodes along the time axis. At t+dt, it is determined whether there is a drivable node and the number of drivable nodes (i.e., child nodes), n(t), is recorded. If a drivable node exists, the costs of n(t) child nodes are calculated and sorted in ascending order. The child node with the smallest cost is selected as the updated parent node, and this process is repeated until the space-time curve is obtained. If so, velocity planning is successful, and the operation terminates. When n(t) is zero, the process backtracks dt time units, selects the child node with the second smallest cost as the updated parent node, and iterates backwards until the space-time curve is obtained. If so, velocity planning is successful, and the operation terminates. If the space-time curve cannot be successfully traversed back dt time units, the third child node with the smallest cost is selected as the updated parent node and the backward recursion operation is executed iteratively. This process is repeated until the velocity planning is successful, and then the operation is terminated.

[0075] Figure 9 A schematic block diagram of an autonomous vehicle speed planning device 900 according to some embodiments of the present disclosure is shown. The device 900, for example, can perform even-numbered... Figure 1The computing device 110 shown may be implemented as a computing device 110.

[0076] like Figure 4 As shown, the device 900 includes a drivable area determination module 901, a node determination module 903, a space-time curve determination module 905, and a target planning speed acquisition module 907. The drivable area determination module 901 is configured to determine the drivable area of ​​the vehicle to be speed-planned in a space-time relationship diagram, which is associated with the vehicle's planned driving trajectory. The node determination module 903 is configured to determine multiple nodes in the space-time relationship diagram, each node indicating the corresponding spatial position of the vehicle in the planned driving trajectory at a specific time. The space-time curve determination module 905 is configured to determine the nodes within the drivable area at the next time moment based on the nodes corresponding to the vehicle at the current time moment, thereby obtaining a space-time curve corresponding to the planned driving trajectory in the space-time relationship diagram. The target planning speed acquisition module 905 is configured to acquire the vehicle's target planning speed at different times based on the space-time curve.

[0077] In some embodiments, the drivable area determination module 901 may also be configured to establish a spatial-temporal relationship diagram based on the planned driving trajectory of the vehicle; project the space and time occupied by one or more obstacles onto the spatial-temporal relationship diagram to obtain the non-drivable area of ​​the vehicle, wherein the predicted driving path of one or more obstacles intersects at least partially with the planned driving trajectory of the vehicle; and determine the area outside the non-drivable area in the spatial-temporal relationship diagram as the drivable area.

[0078] In some embodiments, the node determination module 903 may also be configured to discretize the spatial-temporal relationship diagram into multiple squares at preset intervals; and to determine the center point of each square in the multiple squares as multiple nodes.

[0079] In some embodiments, the space-time curve determination module can also be configured to select a node indicating the spatial position of the vehicle at the current moment as the parent node and recursively push to the next moment; calculate the cost of the drivable child node at the next moment according to the cost function; select the child node with the minimum cost as the updated parent node and continue to recursively push to the subsequent moments of the next moment; and in response to the successful determination of the recursion between the nodes at each moment, determine a series of child nodes as a space-time curve by connecting them.

[0080] In some embodiments, the space-time curve determination module 905 may also be configured to determine the vehicle's comfort constraints and road speed limits, the comfort constraints being associated with one or more of the vehicle's longitudinal velocity, acceleration, jerk, and lateral acceleration; in response to determining that each child node recursively from the updated parent node does not satisfy either the comfort constraints or the road speed limits, the second least costly child node is reselected as the updated parent node for recursion to subsequent time points; and in response to determining that each child node recursively from the second least costly child node as the updated parent node for recursion to subsequent time points does not satisfy either the comfort constraints or the road speed limits, the third least costly child node is reselected as the updated parent node for recursion to subsequent time points.

[0081] In some embodiments, the space-time curve determination module 907 can also be configured to traverse the current moment from the parent node to each child node, and if none of them satisfy either the comfort constraint or the road speed limit, backtrack to the previous moment; and select the child node with the second smallest cost at the previous moment to recursively push to subsequent moments until the recursed node satisfies both the comfort constraint and the road speed limit.

[0082] In some embodiments, the target planning speed acquisition module 907 may also be configured to determine the slope of each node on the space-time curve as the target planning speed, optionally or additionally, to perform secondary smoothing on the space-time curve to calculate the speed of each node on the space-time curve, and to bind the speed of each node to the corresponding trajectory point on the planned driving trajectory to obtain the target planning speed.

[0083] It should be understood that each module described in device 900 is related to a reference. Figure 1 Each step in the described method 100 corresponds to this. Therefore, the above is combined with... Figure 1 The described operations and features also apply to the device 900 and the modules contained therein, and have the same effect; further details will not be elaborated here.

[0084] If speed planning is successful, the method 100 or device 900 selects the child node with the smallest cost as the parent node for each time point in the recursive process and continues recursively. If speed planning fails, nodes with costs from largest to smallest are selected in turn according to the cost value and the recursive process is restarted. If all child nodes at a given time point fail to be recursively selected and the recursive process cannot continue, the process is backtracked to the previous time point and nodes with costs from largest to smallest are selected in turn according to the cost value and the recursive process is restarted until the recursive process is successful.

[0085] Compared with other speed planning methods or devices, this speed planning method 100 or device 900 only considers the availability of nodes at a specific moment, and the overall recursive direction can be carried out in chronological order. It has a small computational load and high efficiency, which can better solve the problem of large computational load and low efficiency in speed planning, reduce the hardware cost of autonomous vehicles, and ensure driving safety and driving experience.

[0086] Figure 10 A schematic block diagram of an example device 1000 that can be used to implement embodiments of the present disclosure is shown. Device 1000 can, for example, be used to implement... Figure 1 The computing device 110.

[0087] like Figure 10 As shown, device 1000 includes a computing unit 1001, which can perform various appropriate actions and processes according to computer program instructions stored in read-only memory (ROM) 1002 or loaded from storage unit 1008 into random access memory (RAM) 1003. The RAM 1003 may also store various programs and data required for the operation of device 1000. The computing unit 1001, ROM 1002, and RAM 1003 are interconnected via bus 1004. Input / output (I / O) interface 1005 is also connected to bus 1004.

[0088] Multiple components in device 1000 are connected to I / O interface 1005, including: input unit 1006, such as keyboard, mouse, etc.; output unit 1007, such as various types of monitors, speakers, etc.; storage unit 1008, such as disk, optical disk, etc.; and communication unit 1009, such as network card, modem, wireless transceiver, etc. Communication unit 1009 allows device 1000 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0089] The computing unit 1001 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 1001 performs the various methods and processes described above, such as method 300. For example, in some embodiments, method 300 may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 1008. In some embodiments, part or all of the computer program may be loaded and / or installed on device 1000 via ROM 1002 and / or communication unit 1009. When the computer program is loaded into RAM 1003 and executed by the computing unit 1001, one or more steps of method 300 described above may be performed. Alternatively, in other embodiments, the computing unit 1001 may be configured to execute method 300 by any other suitable means (e.g., by means of firmware).

[0090] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload programmable logic devices (CPLDs), and so on.

[0091] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0092] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0093] Furthermore, although the operations are described in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, although several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations.

[0094] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A speed planning method for autonomous vehicles, characterized in that, include: The drivable area of ​​the vehicle to be speed-planned is determined in the space-time relationship diagram, which is associated with the planned driving trajectory of the vehicle. Multiple nodes are determined in the space-time relationship diagram, and each of the multiple nodes indicates the corresponding spatial position of the vehicle at a specific time in the planned driving trajectory; Based on the node corresponding to the vehicle at the current moment, determine the node in the drivable area at the next moment, so as to obtain the space-time curve corresponding to the planned driving trajectory in the space-time relationship diagram; as well as Based on the space-time curve, the target planned speed of the vehicle at different times is obtained; The process of determining the node within the drivable area at the next moment based on the node corresponding to the vehicle at the current moment, in order to obtain the space-time curve corresponding to the planned driving trajectory in the space-time relationship diagram, includes: Select the node that indicates the spatial position of the vehicle at the current time as the parent node and recursively push forward to the next time. The cost of the drivable sub-node at the next moment is calculated based on the cost function; The child node with the lowest cost is selected as the updated parent node and the process continues to recursively advance to subsequent times of the next time. When recursively advancing, only the feasible nodes at the corresponding time are considered, and all nodes in the spatial-temporal relationship graph are not considered. In response to the successful recursion between each time node, a series of child nodes are connected to form the space-time curve. If the child node with the lowest cost at the current time fails to be recursively selected, the child node with the second lowest cost is selected as the new parent node for recursion. If the child node with the second lowest cost still cannot be recursively selected, the child node with the third lowest cost is selected as the parent node for recursion and the above steps are repeated. If all child nodes at the current time fail to be recursively selected, the process is backtracked to the previous time, the child node with the second lowest cost at the previous time is selected again for recursion, and the above steps are repeated.

2. The method according to claim 1, characterized in that, The drivable area of ​​the vehicle to be speed-planned is determined in the space-time relationship diagram, including: Based on the planned driving trajectory of the vehicle, the spatial-temporal relationship diagram is established; Projecting the space and time occupied by one or more obstacles onto the space-time relationship diagram yields the drivable area of ​​the vehicle, wherein the predicted driving paths of one or more of the obstacles at least partially intersect the planned driving trajectory of the vehicle; and The area outside the non-drivable area in the space-time relationship diagram is defined as the drivable area.

3. The method according to claim 1, characterized in that, The multiple nodes identified in the space-time relationship diagram include: The spatial-temporal relationship diagram is discretized into multiple squares at preset intervals; and The center point of each of the multiple squares is determined as one of the multiple nodes.

4. The method according to claim 1, characterized in that, Selecting the child node with the lowest cost as the updated parent node and continuing the process recursively to subsequent time steps includes: Determine the comfort constraints and road speed limits for the vehicle, wherein the comfort constraints are associated with one or more of the vehicle’s longitudinal velocity, acceleration, jerk, and lateral acceleration. In response to the determination that none of the child nodes recursively from the updated parent node satisfy either the comfort constraint or the road speed limit, the child node with the second smallest cost is reselected as the updated parent node for subsequent time steps; and In response to the fact that each child node determined to be the second least costly child node and subsequently recursively assigned to it does not satisfy either the comfort constraint or the road speed limit, the third least costly child node is reselected as the updated parent node and recursively assigned to it for subsequent time steps.

5. The method according to claim 4, characterized in that, Selecting the child node with the lowest cost as the updated parent node and continuing the process recursively to subsequent time steps includes: If, during the current moment, recursively traversing from the parent node to each child node, neither the comfort constraint nor the road speed limit is satisfied, then backtrack to the previous moment; and Select the child node with the second smallest cost at the previous time step and recursively push forward to subsequent time steps until the recursed node satisfies the comfort constraint and the road speed limit.

6. The method according to any one of claims 1 to 5, wherein obtaining the target planned speed of the vehicle at different times based on the space-time curve comprises: The slope of each node on the space-time curve is determined as the target planning speed; and / or The space-time curve is smoothed twice to calculate the velocity of each node on the space-time curve, and the velocity of each node is bound to the corresponding trajectory point on the planned driving trajectory to obtain the target planned velocity.

7. A speed planning device for an autonomous vehicle, characterized in that, include: The drivable area determination module is configured to determine the drivable area of ​​the vehicle to be speed-planned in a space-time relationship diagram, wherein the space-time relationship diagram is associated with the planned driving trajectory of the vehicle. A node determination module is configured to determine multiple nodes in the space-time relationship diagram, each of the multiple nodes indicating the corresponding spatial position of the vehicle at a specific time in the planned driving trajectory; The space-time curve determination module is configured to determine the node in the drivable area at the next moment based on the node corresponding to the vehicle at the current moment, so as to obtain the space-time curve corresponding to the planned driving trajectory in the space-time relationship diagram. as well as The target planning speed acquisition module is configured to acquire the target planning speed of the vehicle at different times based on the space-time curve; The process of determining the node within the drivable area at the next moment based on the node corresponding to the vehicle at the current moment, in order to obtain the space-time curve corresponding to the planned driving trajectory in the space-time relationship diagram, includes: Select the node that indicates the spatial position of the vehicle at the current time as the parent node and recursively push forward to the next time. The cost of the drivable sub-node at the next moment is calculated based on the cost function; The child node with the lowest cost is selected as the updated parent node and the process continues to recursively advance to subsequent times of the next time. When recursively advancing, only the feasible nodes at the corresponding time are considered, and all nodes in the spatial-temporal relationship graph are not considered. In response to the successful recursion between each time node, a series of child nodes are connected to form the space-time curve. If the child node with the lowest cost at the current time fails to be recursively selected, the child node with the second lowest cost is selected as the new parent node for recursion. If the child node with the second lowest cost still cannot be recursively selected, the child node with the third lowest cost is selected as the parent node for recursion and the above steps are repeated. If all child nodes at the current time fail to be recursively selected, the process is backtracked to the previous time, the child node with the second lowest cost at the previous time is selected again for recursion, and the above steps are repeated.

8. An electronic device, the device comprising: One or more processors; as well as A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, the program, when executed by a processor, implementing the method according to any one of claims 1 to 6.

Images