Vehicle longitudinal planning method, computer program product, device and storage medium

Abstract

The present application relates to the field of autonomous driving technology, and more specifically, to a vehicle longitudinal planning method, computer program product, device, and storage medium. The method includes performing scene recognition on the driving environment surrounding an autonomous vehicle to determine the complexity of the current scene; determining a matching longitudinal planning model based on the complexity of the current scene; obtaining target longitudinal driving parameters of the autonomous vehicle based on the longitudinal planning model; and performing longitudinal control of the autonomous vehicle based on the target longitudinal driving parameters. The present application can match corresponding longitudinal planning models in both simple and complex scenarios, thereby improving the adaptability of the scenario and the flexibility of autonomous driving.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and more specifically, to a vehicle longitudinal planning method, computer program product, device, and storage medium. Background Art

[0002] In autonomous driving systems, longitudinal speed planning (LSP) is a core module for vehicle decision-making and control. Its primary goal is to determine the vehicle's longitudinal speed, acceleration, and corresponding braking / acceleration actions based on the surrounding traffic environment (such as preceding vehicles, pedestrians, traffic signs / lights, and intersections) to achieve safe, comfortable, and efficient driving.

[0003] In the existing technology, the longitudinal speed of a vehicle is mainly controlled based on information such as the relative distance and relative speed between the vehicle itself and the vehicle in front. This method results in insufficient adaptability and flexibility of the vehicle to complex scenarios.

[0004] In view of this, this application is filed. Summary of the Invention

[0005] The purpose of this application is to provide a vehicle longitudinal planning method, computer program product, device and storage medium to adapt to driving environments in different scenarios and improve driving flexibility.

[0006] In order to achieve the above objectives, this application adopts the following technical solutions:

[0007] In a first aspect, the present application provides a vehicle longitudinal planning method, comprising:

[0008] Perform scene recognition on the vehicle's surrounding driving environment to obtain the complexity of the current scene;

[0009] Determining a matching longitudinal planning mode according to the complexity of the current scenario;

[0010] obtaining target longitudinal driving parameters of the vehicle according to the longitudinal planning model;

[0011] The vehicle is longitudinally controlled according to the target longitudinal driving parameter.

[0012] In a second aspect, the present application provides a computer program product, comprising:

[0013] The computer program product stores computer instructions, which implement the steps of the vehicle longitudinal planning method when executed by a processor.

[0014] In a third aspect, the present application provides an electronic device, comprising:

[0015] at least one processor, and a memory communicatively coupled to the at least one processor;

[0016] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform a vehicle longitudinal planning method.

[0017] In a fourth aspect, the present application provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to enable the computer to execute the above-mentioned vehicle longitudinal planning method.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] This application performs scene recognition on the surrounding driving environment of the autonomous driving vehicle to obtain the complexity of the current scene. The matching longitudinal planning mode can be determined based on the complexity of the current scene, so that the corresponding longitudinal planning mode can be matched in both simple and complex scenes, thereby improving the adaptability of the scene and the flexibility of autonomous driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the specific implementation methods or the description of the prior art. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 This is a flow chart of a vehicle longitudinal planning method provided in an embodiment of the present application;

[0022] Figure 2 It is a structural diagram of the electronic device provided in this application. DETAILED DESCRIPTION

[0023] The following description of exemplary embodiments of the present application is made in conjunction with the accompanying drawings, including various details of the embodiments of the present application to facilitate understanding. These details should be considered as merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present application. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0024] As mentioned in the background art, the existing vehicles have insufficient adaptability and flexibility to complex scenarios. The present application will be further described in detail below with reference to the embodiments.

[0025] Figure 1This is a flow chart of a vehicle longitudinal planning method provided in this embodiment. This embodiment is applicable to situations where longitudinal planning of an autonomous vehicle is performed, such as determining the longitudinal speed, acceleration, and corresponding braking / acceleration actions of the autonomous vehicle to achieve safe, comfortable, and efficient driving.

[0026] like Figure 1 As shown, this embodiment provides a vehicle longitudinal planning method, comprising the following steps:

[0027] S110: Perform scene recognition on the driving environment surrounding the autonomous driving vehicle to obtain the complexity of the current scene.

[0028] Optionally, the autonomous vehicle's perception system and / or vehicle-road-cloud collaborative technology can be used to identify surrounding vehicles, obstacles, road structure, and traffic light status in the driving environment to determine the current scene. Vehicle perception devices include, but are not limited to, radar and cameras. Road structures include, but are not limited to, highways, urban roads, ramps, and intersections.

[0029] Optionally, sense the longitudinal position of surrounding vehicles (such as the vehicle in front) , horizontal , longitudinal speed , lateral velocity , longitudinal acceleration , lateral acceleration , and the location information of the autonomous vehicle; a behavior prediction model based on a constant acceleration model (see the following formula) or reinforcement learning (RL) training is used to predict the target vehicle’s future (2~4 seconds) trajectory:

[0030] ;

[0031] ;

[0032] , Indicates the value range of the prediction time domain.

[0033] Behavior prediction models trained with reinforcement learning can be used to predict the trajectories of surrounding vehicles, thereby improving the robustness and safety of longitudinal control. Optionally, a state space is constructed using the driving states of surrounding vehicles, the autonomous vehicle, and environmental information. The action space is then defined by defining the possible behaviors of surrounding vehicles. The design goal is to ensure that the predicted trajectories closely match actual driving behavior. Commonly used reward criteria include trajectory similarity, physical feasibility, and interaction rationality.

[0034] Next, based on the lateral speed or acceleration of the surrounding vehicles in the predicted time domain, determine whether the surrounding vehicles are cutting in or out; and identify whether it is an intersection, ramp entrance or exit based on the map or lane structure.

[0035] The scenarios in this embodiment include, but are not limited to, entry and exit scenarios, intersection scenarios, on-ramp and off-ramp scenarios, vehicle following scenarios, and cruising scenarios. Different scenarios have varying degrees of complexity. For example, vehicle following scenarios require focusing on the vehicle ahead and paying less attention to the driving conditions of surrounding vehicles, thus resulting in a lower level of complexity. Entry and exit scenarios require attention to information such as the direction and speed of surrounding vehicles, as well as lane width, thus resulting in a higher level of complexity. Those skilled in the art may assign a complexity variable to each scenario.

[0036] S120: Determine a matching longitudinal planning mode based on the complexity of the current scene.

[0037] S130. Obtain target longitudinal driving parameters of the autonomous driving vehicle according to the longitudinal planning model.

[0038] The longitudinal planning mode determines the target longitudinal driving parameters of the autonomous vehicle, such as longitudinal acceleration and longitudinal velocity. Optional longitudinal planning modes include kinematics-based longitudinal planning, interpolation-based multi-objective velocity planning, adaptive PID (proportional-derivative-integral) with torque compensation, model predictive control, dynamic programming, and quadratic programming.

[0039] Different longitudinal planning modes are suitable for scenarios of different complexity. When selecting a matching longitudinal planning mode, you can match it based on the difficulty of solving each longitudinal planning mode and the accuracy of longitudinal control in scenarios of different complexity. For example, in the less complex following and cruising scenarios, the kinematics-based longitudinal planning mode has a lower difficulty of solving and a higher accuracy, so the kinematics-based longitudinal planning mode is matched. For example, in the more complex cut-in and cut-out scenarios, the kinematics-based longitudinal planning mode has a low accuracy, while the dynamic planning mode has a high accuracy, so the dynamic planning mode is matched.

[0040] S140. Perform longitudinal control on the autonomous driving vehicle according to the target longitudinal driving parameters.

[0041] The vehicle controller drives or brakes the autonomous vehicle based on the target longitudinal driving parameter to achieve the target longitudinal driving parameter. For example, if the target longitudinal driving parameter is a longitudinal speed of 50 km / h and the current speed is 30 km / h, the vehicle will be driven to achieve an acceleration of 50 km / h.

[0042] This application performs scene recognition on the surrounding driving environment of the autonomous driving vehicle to obtain the complexity of the current scene. The matching longitudinal planning mode can be determined based on the complexity of the current scene, so that the corresponding longitudinal planning mode can be matched in both simple and complex scenes, thereby improving the adaptability of the scene and the flexibility of autonomous driving.

[0043] Optionally, the scenarios are divided into a first category and a second category, where the complexity of the first category is less than that of the second category. If the current scenario belongs to the first category, a first longitudinal planning mode based on kinematics is determined. If the current scenario belongs to the second category, a second longitudinal planning mode is determined. The second longitudinal planning mode includes at least one of a distance-time ST graph, dynamic programming (DP), and quadratic programming (QP).

[0044] Among them, the first longitudinal planning mode based on kinematics is mainly based on information such as the relative distance and relative speed between the vehicle itself and the vehicle in front, and performs longitudinal control through kinematic formulas and preset logical rules (such as time distance maintenance, minimum safety distance constraints, speed limits, etc.).

[0045] In some simple scenarios, the first longitudinal planning mode is required to obtain the target longitudinal driving parameters of the autonomous driving vehicle. The advantages of the first longitudinal planning mode include: 1) Good stability: Since the formulas and rules are relatively fixed, the overall output is stable and the sensitivity to sensory noise is low; 2) Simple implementation: The algorithm implementation logic is relatively concise and occupies less hardware resources; 3) High reliability: It performs consistently and safely in simple scenarios such as structured roads and cruise control. The disadvantages of the first longitudinal planning mode include: 1) Insufficient adaptability to complex scenarios: In dynamic scenarios such as urban roads, frequent lane changes, and congested intersections, this mode is often more passive in its response and cannot fully utilize the predicted information of the surrounding target trajectories, resulting in low accuracy; 2) Insufficient flexibility: It is unable to make fast and optimal decision-making plans for sudden scenarios (such as sudden braking and sudden cutting) in a short period of time.

[0046] In some complex scenarios, a second longitudinal planning model is needed to determine the target longitudinal driving parameters of the autonomous vehicle. For example, longitudinal planning models based on ST graphs, dynamic programming, and quadratic programming have the following advantages: 1) Strong adaptability to complex scenarios: They can flexibly handle urban intersections, on- and off-ramps, and situations where the target vehicle frequently cuts in and out; 2) Utilize short-term predictions: Trajectory predictions are based on information such as the speed and acceleration of surrounding vehicles, making decisions more forward-looking. However, their disadvantages include: 1) Dependence on perception accuracy: Inaccurate estimates of target speed or acceleration can result in significant discrepancies between the predicted and actual trajectories, thus affecting planning effectiveness; 2) High solution complexity: In real-time systems, high computational performance is required. If the algorithm fails to solve in extreme scenarios or is too time-consuming, it can cause safety or user experience issues; 3) Potentially weak consistency: Different environmental noise, perception errors, or differences in future trajectory assumptions can significantly fluctuate planning results, resulting in an inconsistent driving experience.

[0047] This embodiment adopts a scenario-driven dual-mode longitudinal planning strategy, which can select different planning algorithms according to the complexity of the scenario, taking into account both the flexibility of complex scenarios and the stability of simple scenarios.

[0048] In some embodiments, the longitudinal planning mode is a first longitudinal planning mode; and for a vehicle-following scenario, obtaining target longitudinal driving parameters of the autonomous driving vehicle according to the longitudinal planning mode includes the following steps:

[0049] The first step is to collect the relative distance between the autonomous vehicle and the vehicle in front if the current scenario is a following vehicle.

[0050] Assume that the expected time distance between the autonomous vehicle and the preceding vehicle is h, then the expected distance for:

[0051] ;

[0052] in, is the speed of the autonomous vehicle.

[0053] Current relative distance for:

[0054] ;

[0055] in, and Represent the positions of the preceding vehicle and the autonomous driving vehicle respectively;

[0056] The second step is to determine the speed correction by minimizing the difference between the relative distance and the expected distance.

[0057] Will As the difference between the relative distance and the expected distance (i.e., position error), it is input into a closed-loop controller (such as a proportional-differential-integral controller) to obtain the speed correction , this speed correction can minimize the position error at the current moment and realize the control of the position loop.

[0058] Step 3: Determine the target speed based on the speed correction , see the following formula:

[0059] ;

[0060] in, is the speed of the autonomous vehicle at the previous moment.

[0061] Step 4: Determine the target longitudinal acceleration by minimizing the difference between the current speed and the target speed.

[0062] Compare the current speed to the target speed The difference is input as an error into the closed-loop controller (such as the proportional-differential-integral controller) to obtain the target longitudinal acceleration. , to realize the control of the speed loop.

[0063] In some embodiments, the longitudinal planning mode is a first longitudinal planning mode; and for a cruising scenario, obtaining target longitudinal driving parameters of the autonomous driving vehicle according to the longitudinal planning mode includes the following steps:

[0064] Step 1: If the current scene is cruise, determine the target speed to be the cruise setting speed .

[0065] The second step is to determine the target longitudinal acceleration by minimizing the difference between the current speed and the target speed.

[0066] Compare the current speed to the target speed The difference is input as an error into the closed-loop controller (such as the proportional-differential-integral controller) to obtain the target longitudinal acceleration. , to realize the control of the speed loop.

[0067] Whether in a cruising scenario or a vehicle-following scenario, the following logical rules must be met when obtaining the target longitudinal acceleration using the first longitudinal planning mode:

[0068] 1) Safety distance check: If , then the emergency brake is triggered immediately;

[0069] 2) Speed limit: (road speed limit or vehicle limit);

[0070] 3) Acceleration limit: .

[0071] This embodiment adopts a reliable solution based on kinematics and logic rules, adopts closed-loop control principles such as position loop and speed loop, and cooperates with safety distance and time interval restrictions to ensure that the vehicle has stable and consistent longitudinal behavior in the absence of special emergency scenarios.

[0072] In some embodiments, the longitudinal planning mode is a second longitudinal planning mode; and obtaining target longitudinal driving parameters of the autonomous driving vehicle according to the longitudinal planning mode comprises the following steps:

[0073] The first step is to construct a distance-time ST graph based on the relative distance between the autonomous driving vehicle and surrounding vehicles or obstacles; and determine the feasible area based on the ST graph.

[0074] The ST diagram is a core tool for longitudinal planning. Its horizontal axis represents time (T) and its vertical axis represents the relative distance (S) between the autonomous vehicle and the vehicle ahead. It describes the vehicle's future motion and is used for collision detection and feasible speed range calculation. The ST diagram's horizontal axis represents time t, while the vertical axis represents the relative distance d between the autonomous vehicle and surrounding vehicles or obstacles. Within the time range [0, T], the motion space of surrounding vehicles or obstacles is predicted and mapped into a series of infeasible regions (obstacle regions). Conversely, the remaining regions represent feasible areas for the autonomous vehicle.

[0075] The second step is to perform dynamic programming search in the feasible area to obtain candidate paths.

[0076] Dynamic programming is one of the core algorithms for solving velocity planning problems in ST graphs. Its basic idea is to find the globally optimal trajectory by constructing a two-dimensional cost table and calculating the cost value for each grid.

[0077] First, the feasible region is discretized into several trajectory points (t i ,d j );in, t i is the i-th moment, d j The cost of each trajectory point is calculated based on the cost function, and a dynamic programming algorithm (such as A* or Dijkstra algorithm) is used to search the cost table to obtain the path with the minimum cost, that is, the rough candidate path.

[0078] Optionally, when constructing the cost function, this embodiment incorporates indicators such as safety distance, collision cost, vehicle speed smoothness, and ride comfort into the cost function. For safety distance, the ST graph is constructed to detect whether the candidate path collides with obstacles and surrounding vehicles. If there is a risk of collision, the corresponding penalty term M is added. For collision cost, the distance between the autonomous driving vehicle and obstacles and surrounding vehicles is included in the cost function. The smaller the distance, the greater the penalty D. For vehicle speed smoothness, acceleration is used as an indicator to measure vehicle speed smoothness. For ride comfort, jerk is used as an indicator to measure ride comfort. Based on the above analysis, the cost function J is constructed as follows:

[0079] ;

[0080] in, a i is the acceleration at the i-th moment, j i is the acceleration at the i-th moment, is the i-th duration, M i is the penalty term for collision risk at the i-th moment, D i is the penalty term of the collision cost at the i-th moment. is the weight coefficient.

[0081] The third step is to optimize the candidate path based on quadratic programming to obtain the target longitudinal acceleration curve.

[0082] While dynamic programming can provide a rough optimal solution, the resulting candidate paths are often not smooth enough. To obtain a smoother control curve, quadratic programming is used to optimize the dynamic programming results. The goal of quadratic programming is to fit a piecewise linear trajectory into a smooth nonlinear curve, thereby improving trajectory continuity and comfort.

[0083] Use the candidate paths found by dynamic programming as the initial solution or constraint region. Construct a quadratic programming problem, such as minimizing acceleration, satisfying speed and acceleration upper and lower limits, and safety distance constraints:

[0084] ;

[0085] in, and is the weight coefficient, is the acceleration at time t, is the acceleration at time t, is the speed at time t, and are the minimum and maximum speed values respectively. and are the minimum and maximum acceleration values, is the distance between the autonomous vehicle and surrounding vehicles or obstacles at time t, For a safe distance.

[0086] Solve the target longitudinal acceleration curve , and then get the optimal speed curve .

[0087] This embodiment is based on the optimization planning of short-term target prediction. In complex scenarios, it uses the driving information of surrounding vehicles and obstacles to construct the ST graph and cost function for dynamic planning and quadratic planning, which can control the autonomous driving vehicle to quickly converge to the optimal trajectory.

[0088] In some schemes, if the target longitudinal driving parameters cannot be obtained according to the second longitudinal planning mode, for example, the solution takes too long, the solution fails, there is no response, or the solver of the second longitudinal planning mode crashes, then the system falls back to the first longitudinal planning mode, and according to the first longitudinal planning mode, the target longitudinal driving parameters of the autonomous driving vehicle are obtained to ensure that the autonomous driving vehicle does not encounter the dangerous situation of no speed command or no acceleration command.

[0089] In some embodiments, longitudinally controlling the autonomous vehicle based on the target longitudinal travel parameter includes smoothing the current target longitudinal travel parameter based on the target longitudinal travel parameter at a historical time when switching longitudinal planning modes or when the target longitudinal travel parameter is updated. This embodiment smoothes or buffers the target longitudinal travel parameter, such as velocity or acceleration, to avoid vehicle jerking or jarring.

[0090] This embodiment does not limit the smoothing method, such as exponential filtering algorithm or S-curve smoothing algorithm. The following formula gives a feasible smoothing example:

[0091] ;

[0092] in, is the weight, t is the time, is the acceleration after smoothing, is the target acceleration, is the acceleration at the previous moment.

[0093] Based on the above embodiments, this application has the following technical effects:

[0094] 1. Adaptability to multiple scenarios: Longitudinal planning mode switching is achieved through scene recognition. In simple scenarios such as highways and following vehicles, the first longitudinal planning mode based on kinematics is used to ensure stability and consistency. In complex scenarios such as urban roads, intersections, entry and exit, the second longitudinal planning mode is used to quickly respond to dynamic changes in the scene.

[0095] 2. Ensure safety and user experience: When a fault or solution failure occurs in the longitudinal planning mode, it can fall back to the first longitudinal planning mode based on kinematics in real time to avoid loss of control or generation of dangerous instructions.

[0096] 3. Reduce reliance on perception uncertainty: For simple scenarios, reduce over-reliance on target acceleration accuracy; only enable high-accuracy predictions in complex scenarios that require short-term predictions, so that the vehicle remains stable most of the time.

[0097] 4. Flexible expansion: The design based on scene recognition and longitudinal planning mode can gradually introduce more environmental information (such as traffic light recognition, pedestrian detection, etc.), and perform safe and efficient longitudinal planning under more complex working conditions.

[0098] like Figure 2 As shown, this embodiment provides an electronic device, including:

[0099] at least one processor; and

[0100] a memory communicatively connected to at least one of the processors; wherein,

[0101] The memory stores instructions executable by at least one processor, and the instructions are executed by at least one processor to enable at least one processor to perform the above method. The at least one processor in the electronic device is capable of performing the above method, thereby having at least the same advantages as the above method.

[0102] Optionally, the electronic device also includes interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. The various components are connected to each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed in the electronic device, including instructions stored in or on the memory to display graphical information of a GUI (Graphical User Interface) on an external input / output device (such as a display device coupled to the interface). In other embodiments, if necessary, multiple processors can be used with multiple memories, and / or multiple buses can be used with multiple memories. Similarly, multiple electronic devices can be connected (for example, as a server array, a group of blade servers, or a multi-processor system), with each device providing part of the necessary operations. Figure 2 A processor 301 is taken as an example.

[0103] Memory 302, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the vehicle longitudinal planning method in the embodiments of this application. Processor 301 executes the software programs, instructions, and modules stored in memory 302 to execute various functional applications and data processing of the device, thereby implementing the aforementioned vehicle longitudinal planning method.

[0104] The memory 302 may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function; the data storage area may store data created based on the use of the terminal, etc. Furthermore, the memory 302 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 302 may further include a memory remotely located relative to the processor 301, and these remote memories may be connected to the device via a network. Examples of the aforementioned networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0105] The electronic device may further include: an input device 303 and an output device 304. The processor 301, the memory 302, the input device 303 and the output device 304 may be connected via a bus or other means. Figure 2 The bus connection is taken as an example.

[0106] The input device 303 can receive input digital or character information, and the output device 304 can include a display device, an auxiliary lighting device (e.g., an LED), a tactile feedback device (e.g., a vibration motor), etc. The display device may include, but is not limited to, a liquid crystal display (LCD), a light-emitting diode (LED) display, and a plasma display. In some embodiments, the display device may be a touch screen.

[0107] This embodiment provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to cause the computer to execute the above-described method. The computer instructions on the computer-readable storage medium are used to cause the computer to execute the above-described method, thereby having at least the same advantages as the above-described method.

[0108] The medium in this application may be any combination of one or more computer-readable media. The medium may be a computer-readable signal medium or a computer-readable storage medium. The medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples of media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device.

[0109] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0110] The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wire, optical cable, RF (Radio Frequency), etc., or any suitable combination of the foregoing.

[0111] The computer program code for performing the operations of the present application can be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0112] In the above embodiments, all or part of the embodiments can be implemented using software, hardware, firmware, or any combination thereof. When implemented using software, all or part of the embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired connection, such as a coaxial cable, optical fiber, digital subscriber line (DSL), or wireless connection, such as infrared, wireless, or microwave. The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data storage device, such as a server or data center, that includes one or more available media. The available medium can be a magnetic medium (e.g., a floppy disk, hard disk, or magnetic tape), an optical medium, or a semiconductor medium. It is worth noting that the computer-readable storage medium mentioned in the embodiments of the present application may be a non-volatile storage medium, in other words, a non-transitory storage medium.

[0113] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this application can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this application can be achieved. This is not a limitation herein.

[0114] The above specific embodiments do not constitute a limitation on the scope of protection of this application. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application shall be included within the scope of protection of this application.

Claims

1. A vehicle longitudinal planning method, characterized in that: include: Perform scene recognition on the driving environment around the autonomous vehicle to obtain the complexity of the current scene; Determining a matching longitudinal planning mode according to the complexity of the current scene, including: if the current scene belongs to a first type of scene, determining a first longitudinal planning mode based on kinematics; if the current scene belongs to a second type of scene, determining a second longitudinal planning mode; wherein the second longitudinal planning mode includes at least one of a distance-time ST graph, dynamic programming, and quadratic programming; wherein the complexity of the first type of scene is less than the complexity of the second type of scene; Among them, the longitudinal planning mode that matches the scenario is selected based on the difficulty of solving each longitudinal planning mode and the accuracy of longitudinal control in scenarios of different complexity. Obtaining target longitudinal travel parameters of the autonomous driving vehicle according to the longitudinal planning mode, including: switching to the first longitudinal planning mode if the target longitudinal travel parameters cannot be obtained according to the second longitudinal planning mode; and obtaining the target longitudinal travel parameters of the autonomous driving vehicle according to the first longitudinal planning mode; The autonomous driving vehicle is longitudinally controlled according to the target longitudinal driving parameter.

2. The method according to claim 1, characterized in that The longitudinal planning mode is a first longitudinal planning mode; Obtaining target longitudinal driving parameters of the autonomous driving vehicle according to the longitudinal planning model includes: If the current scenario is a following vehicle, the relative distance between the autonomous vehicle and the vehicle in front is collected; determining a speed correction by minimizing the difference between the relative distance and the desired distance; determining a target speed according to the speed correction amount; The target longitudinal acceleration is determined by minimizing the difference between the current speed and the target speed.

3. The method according to claim 1, characterized in that The longitudinal planning mode is a first longitudinal planning mode; Obtaining target longitudinal driving parameters of the autonomous driving vehicle according to the longitudinal planning model includes: If the current scene is cruising, the target speed is determined to be the cruise setting speed; The target longitudinal acceleration is determined by minimizing the difference between the current speed and the target speed.

4. The method according to claim 1, wherein The longitudinal planning mode is the second longitudinal planning mode; Obtaining target longitudinal driving parameters of the autonomous driving vehicle according to the longitudinal planning model includes: Constructing a distance-time ST graph based on the relative distances between the autonomous driving vehicle and surrounding vehicles or obstacles; and determining a feasible area based on the ST graph; Perform dynamic programming search in the feasible area to obtain candidate paths; The candidate paths are optimized based on quadratic programming to obtain a target longitudinal acceleration curve.

5. The method according to any one of claims 1 to 4, characterized in that The method further includes: performing longitudinal control of the autonomous driving vehicle according to the target longitudinal driving parameter, comprising: When the longitudinal planning mode is switched or the target longitudinal driving parameter is updated, the target longitudinal driving parameter at the current moment is smoothed based on the target longitudinal driving parameter at the historical moment.

6. A computer program product, characterized in that include: The computer program product stores computer instructions, which, when executed by a processor, implement the steps of the vehicle longitudinal planning method according to any one of claims 1 to 5.

7. An electronic device, characterized in that: include: at least one processor, and a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor to enable at least one processor to perform the vehicle longitudinal planning method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that The medium stores computer instructions, and the computer instructions are used to enable the computer to execute the vehicle longitudinal planning method according to any one of claims 1 to 5.

Images