Longitudinal speed trajectory generation method and electronic device

Abstract

The application discloses a longitudinal speed trajectory generation method and an electronic device. The method comprises the following steps: determining an obstacle object included in a driving environment of a self-driving object; determining a first obstacle object as a reference obstacle object when a corresponding first prediction distance is less than a prediction safety distance and a corresponding first timestamp is earlier than a first reference timestamp; generating a reference longitudinal speed trajectory based on a current driving state of the self-driving object and a predicted driving state of the self-driving object at the first timestamp; and updating the reference longitudinal speed trajectory until a target longitudinal speed trajectory meeting a target condition is obtained, in a case that a corresponding second prediction distance is less than the prediction safety distance and a corresponding second timestamp is earlier than a second reference timestamp based on the reference longitudinal speed trajectory. The application solves the technical problem of low efficiency of vehicle longitudinal speed generation.

Description

Technical Field

[0001] This application relates to the field of computers, and more specifically, to a method for generating longitudinal velocity trajectories and an electronic device. Background Technology

[0002] When autonomous vehicles change lanes, they need to plan their longitudinal speed in real time to ensure safety and smoothness. Existing technologies for planning longitudinal speed during lane changes typically use information such as the distance between the vehicle and all nearby obstacles, as well as defined functions and constraints, to find the optimal longitudinal speed trajectory.

[0003] However, the longitudinal speed generation method provided by the above-mentioned technologies requires a lot of time for large-scale data processing and complex calculation iterations when vehicles change lanes, resulting in a technical problem of low efficiency in generating longitudinal speed.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This application provides a method and electronic device for generating longitudinal velocity trajectories, which at least solves the technical problem of low efficiency in generating longitudinal trajectories.

[0006] According to one aspect of the embodiments of this application, a method for generating a longitudinal speed trajectory is provided, comprising: determining obstacle objects included in the driving environment of a vehicle object; determining a first obstacle object whose corresponding first predicted distance is less than a predicted safe distance and whose corresponding first timestamp is earlier than a first reference timestamp as a reference obstacle object, wherein the first predicted distance is the predicted distance between the vehicle object and the first obstacle object obtained at the first timestamp; generating a reference longitudinal speed trajectory based on the current driving state of the vehicle object and the predicted driving state of the vehicle object at the first timestamp, wherein the predicted driving state is the driving state that the vehicle object needs to achieve at the first timestamp in order to avoid the first obstacle object, and when the vehicle object drives according to the reference longitudinal speed trajectory at the first timestamp, the predicted distance between it and the first obstacle object will be greater than the predicted safe distance; updating the reference longitudinal speed trajectory when a second obstacle object is predicted based on the reference longitudinal speed trajectory to have a corresponding second predicted distance less than the predicted safe distance and a corresponding second timestamp earlier than the second reference timestamp, until a target longitudinal speed trajectory that meets the target conditions is obtained, wherein when the vehicle object drives according to the target longitudinal speed trajectory within a target time period, the predicted distance between it and any obstacle object at any timestamp is greater than the predicted safe distance.

[0007] According to another aspect of the embodiments of this application, a longitudinal velocity trajectory generation apparatus is also provided, comprising: a first determining unit, configured to determine obstacle objects included in the driving environment of a vehicle object; a second determining unit, configured to determine a first obstacle object whose corresponding first predicted distance is less than a predicted safe distance and whose corresponding first timestamp is earlier than a first reference timestamp as a reference obstacle object, wherein the first predicted distance is the predicted distance between the vehicle object and the first obstacle object obtained at the first timestamp; and a generation unit, configured to generate a reference longitudinal velocity trajectory based on the current driving state of the vehicle object and the predicted driving state of the vehicle object at the first timestamp, wherein the predicted driving state is the predicted driving state of the vehicle object in the current driving state of the vehicle object and the predicted driving state of the vehicle object at the first timestamp. At the first timestamp, in order to avoid the first obstacle, the vehicle is traveling according to the reference longitudinal speed trajectory at the first timestamp, and the predicted distance between it and the first obstacle will be greater than the predicted safe distance. The update unit is used to update the reference longitudinal speed trajectory when the second predicted distance is less than the predicted safe distance based on the reference longitudinal speed trajectory, and the second timestamp of the corresponding obstacle is earlier than the second reference timestamp of the second obstacle, until the target longitudinal speed trajectory that meets the target conditions is obtained. In this case, when the vehicle is traveling according to the target longitudinal speed trajectory within the target time period, the predicted distance between it and any obstacle at any timestamp is greater than the predicted safe distance.

[0008] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer program, which is configured to execute the above-described longitudinal velocity trajectory generation method at runtime.

[0009] According to another aspect of the embodiments of this application, a computer program product or computer program is provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the longitudinal velocity trajectory generation method as described above.

[0010] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to execute the above-described longitudinal velocity trajectory generation method through the computer program.

[0011] In this embodiment, obstacles in the driving environment of the vehicle are identified. First obstacles whose first predicted distance is less than the predicted safe distance and whose first timestamp is earlier than the first reference timestamp are identified as reference obstacles. The first predicted distance is the predicted distance between the vehicle and the first obstacle obtained at the first timestamp. A reference longitudinal speed trajectory is generated based on the vehicle's current driving state and its predicted driving state at the first timestamp. The predicted driving state is the driving state the vehicle needs to achieve at the first timestamp to avoid the first obstacle. When the vehicle travels according to the reference longitudinal speed trajectory at the first timestamp, the predicted distance between it and the first obstacle will be greater than the predicted safe distance. If, based on the reference longitudinal speed trajectory, a second obstacle is predicted whose second predicted distance is less than the predicted safe distance and whose second timestamp is earlier than the second reference timestamp, the reference longitudinal speed trajectory is updated until a target longitudinal speed trajectory that meets the target conditions is obtained. When the vehicle travels according to the target longitudinal speed trajectory within the target time period, the predicted distance between it and any obstacle at any timestamp is greater than the predicted safe distance. In other words, by adopting the embodiments of this application, the technical effect of improving the generation efficiency of longitudinal velocity is achieved, and the technical problem of low generation efficiency of longitudinal velocity is solved. Attached Figure Description

[0012] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of an optional longitudinal velocity trajectory generation method according to an embodiment of this application; Figure 2 This is a schematic diagram of an optional longitudinal velocity trajectory generation method according to an embodiment of this application; Figure 3 This is a schematic diagram of another optional longitudinal velocity trajectory generation method according to an embodiment of this application; Figure 4 This is a schematic diagram of another optional longitudinal velocity trajectory generation method according to an embodiment of this application; Figure 5 This is a schematic diagram of another optional longitudinal velocity trajectory generation method according to an embodiment of this application; Figure 6 This is a schematic diagram of another optional longitudinal velocity trajectory generation method according to an embodiment of this application; Figure 7 This is a schematic diagram of an optional longitudinal velocity trajectory generation device according to an embodiment of this application; Figure 8 This is a schematic diagram of the structure of an optional electronic device according to an embodiment of this application. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0014] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0015] As an alternative solution, such as Figure 1 As shown, the above-mentioned method for generating longitudinal velocity trajectories includes: S102, determine the obstacle objects included in the driving environment of the vehicle object.

[0016] S104, the first obstacle object whose corresponding first predicted distance is less than the predicted safe distance and whose corresponding first timestamp is earlier than the first reference timestamp is determined as a reference obstacle object, wherein the first predicted distance is the predicted distance between the vehicle object and the first obstacle object obtained at the first timestamp.

[0017] S106. Based on the current driving state of the vehicle object and the predicted driving state of the vehicle object at the first time stamp, a reference longitudinal velocity trajectory is generated. The predicted driving state is the driving state that the vehicle object needs to achieve at the first time stamp in order to avoid the first obstacle object. When the vehicle object is driving according to the reference longitudinal velocity trajectory at the first time stamp, the predicted distance between it and the first obstacle object will be greater than the predicted safe distance.

[0018] S108, if the second predicted distance is less than the predicted safe distance and the second time stamp is earlier than the second reference time stamp of the second obstacle object based on the reference longitudinal velocity trajectory, the reference longitudinal velocity trajectory is updated until the target longitudinal velocity trajectory that meets the target conditions is obtained. Wherein, when the vehicle object travels according to the target longitudinal velocity trajectory within the target time period, the predicted distance between it and any obstacle object at any time stamp is greater than the predicted safe distance.

[0019] It should be noted that the above-mentioned longitudinal velocity trajectory generation method can be applied, but is not limited to, to lane-changing scenarios of vehicle objects, including the generation of longitudinal velocity of vehicle objects before and during lane changes.

[0020] Optionally, the aforementioned autonomous vehicle object may, but is not limited to, refer to a vehicle performing autonomous driving, which is equipped with the necessary sensors and computing units for perceiving the environment, planning paths, and controlling vehicle movement.

[0021] It should be noted that the above-mentioned driving environment may include, but is not limited to, the specific road conditions, surrounding traffic conditions, weather conditions, etc., in the area surrounding the vehicle during its driving process (such as within the predetermined range), and is a set of external factors that need to be considered when the vehicle performs speed planning.

[0022] Furthermore, the aforementioned obstacle may refer to entities that may collide with the vehicle during lane changing or driving, including but not limited to other vehicles and static obstacles.

[0023] Optionally, after determining the obstacle objects included in the driving environment where the vehicle object is located, it may include, but is not limited to: within the current detection period, such as the time period N seconds after the current initial timestamp, performing collision detection between each obstacle object and the vehicle object according to a fixed time interval sequence, wherein the above time sequence includes multiple timestamps, and collision detection between the vehicle object and each of the above obstacle objects needs to be performed at each timestamp.

[0024] Furthermore, the above collision detection can be performed, but is not limited to, the following steps: S1, based on the lateral trajectory of the vehicle and the predicted trajectory of surrounding obstacles within a certain time period, predicts the position of the vehicle at each timestamp in the time series within that period, as well as the position of the obstacles at each timestamp in the time series within that period.

[0025] S2: Using the position of the vehicle at each timestamp in the time series within this period, and the position of the obstacle at each timestamp in the time series within this period, calculate the predicted distance between the vehicle and each obstacle at each timestamp. This predicted distance includes both lateral and longitudinal predicted distances. It should be noted that in this step, if the longitudinal velocity trajectory has already been iteratively obtained, the above steps are performed based on the currently planned longitudinal velocity trajectory. If the longitudinal velocity trajectory has not yet been iteratively obtained, the above steps are performed based on the initial longitudinal velocity trajectory, which is the initial lane-changing trajectory planned without considering any obstacles.

[0026] S3 compares the predicted distance between the vehicle and each obstacle at each timestamp with the safe predicted distance (including lateral and longitudinal safe predicted distances) between each obstacle and the vehicle at each timestamp. The safe predicted distance for each obstacle differs at different timestamps and is proportional to the time difference between the current and initial timestamps and the lateral and longitudinal velocities of the obstacle at the current timestamp. Specifically, the generation process of the safe predicted distance for a specific obstacle (e.g., the first obstacle) at a certain timestamp (e.g., the first timestamp) is illustrated as follows:

[0027]

[0028] in, This represents the longitudinal safety distance threshold (used to illustrate the example of longitudinal safety prediction mentioned above). The lateral safety distance thresholds (used to illustrate the lateral safety prediction example above) are each defined by a baseline value. and a deviation value Composition. The baseline value is a fixed safety value by default. The deviation value takes into account the prediction time (i.e., the time difference between the current timestamp and the initial timestamp) and the lateral and longitudinal velocities of the obstacle at the prediction time. The maximum value of the deviation value is limited to the expected distance through function design. The deviation value is proportional to the prediction time and the lateral and longitudinal velocities mentioned above.

[0029] S4. If the comparison result shows that the predicted distance between the vehicle and an obstacle at a certain timestamp is less than the predicted safe distance, then the obstacle is identified as an obstacle that may collide at that timestamp. Specifically, if the lateral distance between the obstacle and the vehicle is less than the lateral safe distance threshold and the longitudinal distance is less than the longitudinal safe distance threshold, then the vehicle is considered to have a collision risk with the obstacle.

[0030] S5: Among the obstacle objects with collision risk, the obstacle object with the earliest detection timestamp is selected as the highest priority first obstacle object, and the collision detection event is recorded. The longitudinal distance s1 of the vehicle object relative to the current moment at the time of collision, and the longitudinal overlap (soverlap) between the vehicle and the corresponding obstacle are also recorded. For example, if... Figure 2 As shown It is the lateral safe distance threshold between vehicle 201 and obstacle vehicle 202. It is the longitudinal safe distance threshold between vehicle 201 and obstacle vehicle 202, and the overlap ratio. This indicates the longitudinal spatial length occupied by vehicle 201 within the obstacle vehicle 202 and its safety boundary area.

[0031] Optionally, the aforementioned first predicted distance may, but is not limited to, represent the expected relative distance between the vehicle and the first obstacle at a preset first time stamp (i.e., the first time stamp), which is jointly determined by the vehicle's current position, the expected speed change, and the obstacle's motion prediction.

[0032] It should be noted that the predicted safe distance may include, but is not limited to, a minimum safe distance threshold set based on vehicle speed, acceleration capability, braking performance, and traffic rules, to ensure that at the predicted time point, the distance between the vehicle and the obstacle is sufficient to avoid a collision.

[0033] Furthermore, the aforementioned first timestamp may, but is not limited to, refer to a specific future point in time, at which the vehicle needs to reach a predetermined driving state in order to avoid the first obstacle.

[0034] Optionally, the aforementioned first reference timestamp may, but is not limited to, be used to represent a time base for comparing whether the collision points between the obstacle object and the vehicle object occur earlier, thereby determining which obstacles need to be prioritized. Specifically, the aforementioned first reference timestamp may, but is not limited to, be used to indicate the time when collision detection is performed based on a time series, and when each obstacle object other than the first obstacle object has a collision risk.

[0035] It should be noted that the above driving status may include, but is not limited to, the actual current position, speed, acceleration, and other information of the vehicle, which are the basic input data for speed planning.

[0036] Furthermore, the aforementioned predicted driving state may, but is not limited to, refer to the target state of the vehicle at a future point in time in order to avoid an obstacle, including the expected position, speed, and acceleration, etc., and these states must meet the vehicle's safety and comfort requirements.

[0037] Optionally, the aforementioned reference longitudinal velocity trajectory may, but is not limited to, represent the ideal longitudinal velocity change curve planned by the vehicle at the first time stamp in order to avoid the first obstacle. This trajectory must ensure that the distance between the vehicle and the first obstacle is always kept above the predicted safe distance.

[0038] It should be noted that the aforementioned second predicted distance may include, but is not limited to, the predicted distance between the vehicle and the second obstacle at another future time point (second timestamp) when the vehicle is traveling along the reference longitudinal speed trajectory.

[0039] Furthermore, the aforementioned second timestamp may, but is not limited to, refer to another point in time after the first timestamp, which may be the point at which the vehicle object collides with the second obstacle object and collision avoidance planning is required.

[0040] Optionally, the aforementioned second reference timestamp may, but is not limited to, be used to represent a second time reference point, and is used to compare with the second timestamp to determine whether the second obstacle object appeared earlier than expected, thereby deciding whether the velocity trajectory needs to be updated. Specifically, referring to the aforementioned first reference timestamp, the aforementioned first reference timestamp may, but is not limited to, be used to indicate the time when all obstacle objects other than the first obstacle object and the second obstacle object have a collision risk after collision detection is performed based on the time series.

[0041] It should be noted that the above target conditions may include, but are not limited to, a series of constraints and performance indicators, which specify the minimum requirements for the target longitudinal velocity trajectory in terms of safety, comfort, efficiency, etc., and serve as the standard for evaluating whether the candidate velocity trajectory is qualified.

[0042] Furthermore, the aforementioned target longitudinal velocity trajectory may, but is not limited to, refer to one or a set of optimal longitudinal velocity change curves that are ultimately determined through continuous optimization and iterative adjustments, under the premise of satisfying all target conditions, to ensure that the vehicle can drive safely, comfortably and efficiently within the target time period, while avoiding collisions with any obstacles.

[0043] It should be noted that, in this embodiment, the longitudinal speed planning strategy for the vehicle in a lane-changing scenario can be determined based on, but is not limited to, the positions of the vehicle and the obstacle. Specifically, the obstacle avoidance strategy mainly considers the relative position and relative longitudinal speed between the vehicle and the obstacle. If the obstacle is currently behind the vehicle, and the obstacle is behind the vehicle at the time of collision after a set distance threshold, the vehicle is considered to prioritize the acceleration strategy. If the obstacle is currently in front of the vehicle, and the obstacle is in front of the vehicle before a set distance threshold at the time of collision, the vehicle is considered to prioritize the deceleration strategy. Otherwise, based on the relative speed between the vehicle and the obstacle at the prediction time, the time required for the vehicle to accelerate and decelerate to avoid the obstacle is calculated, and the strategy with the shorter avoidance time is selected as the obstacle avoidance strategy. This embodiment, by prioritizing the determination of the optimal avoidance strategy based on the motion information of surrounding vehicles and the vehicle, provides a better range of solutions for the subsequent generation of the speed polynomial curve trajectory, enabling the algorithm to output the best-performing speed trajectory more quickly, greatly improving computational efficiency. In other words, before the lane change, if... Figure 4 If obstacle 402 is ahead of the vehicle 401, a deceleration strategy will be prioritized. Figure 5 If obstacle object 502 is behind vehicle object 501, then an acceleration strategy is prioritized. Figure 6 If the obstacle object 602 and the vehicle object 601 are relatively parallel, then the strategy with the shortest time consumption is selected from the acceleration strategy and the deceleration strategy.

[0044] In this embodiment, obstacles in the driving environment of the vehicle are identified. First obstacles whose first predicted distance is less than the predicted safe distance and whose first timestamp is earlier than the first reference timestamp are identified as reference obstacles. The first predicted distance is the predicted distance between the vehicle and the first obstacle obtained at the first timestamp. A reference longitudinal speed trajectory is generated based on the vehicle's current driving state and its predicted driving state at the first timestamp. The predicted driving state is the driving state the vehicle needs to achieve at the first timestamp to avoid the first obstacle. When the vehicle travels according to the reference longitudinal speed trajectory at the first timestamp, the predicted distance between it and the first obstacle will be greater than the predicted safe distance. If, based on the reference longitudinal speed trajectory, a second obstacle is predicted whose second predicted distance is less than the predicted safe distance and whose second timestamp is earlier than the second reference timestamp, the reference longitudinal speed trajectory is updated until a target longitudinal speed trajectory that meets the target conditions is obtained. When the vehicle travels according to the target longitudinal speed trajectory within the target time period, the predicted distance between it and any obstacle at any timestamp is greater than the predicted safe distance. By employing the embodiments of this application, obstacles in the driving environment are identified, and based on the predicted distance and timestamp from the vehicle, the earliest collision risk is determined from which obstacle is designated as the "reference obstacle." This allows for focusing on the most pressing collision risks, rather than treating all obstacles equally, thereby reducing computational load and improving efficiency. In lane-changing scenarios, prioritizing the nearest collision risk ensures the vehicle can react promptly and avoid immediate danger. Based on the current and predicted driving states, a "reference longitudinal velocity trajectory" is generated, aiming to allow the vehicle to reach a safe position at the first timestamp to avoid the reference obstacle. This velocity trajectory generation strategy considers not only the vehicle's immediate state but also the needs of future states, ensuring the planned velocity trajectory effectively avoids obstacles. It also takes into account acceleration and jerk limitations to ensure driving comfort and safety. Once a new obstacle (the second obstacle) is predicted to potentially cause a collision, the original reference longitudinal velocity trajectory is immediately updated to ensure the vehicle can complete the lane change while maintaining a predicted safe distance from any obstacle. This approach improves the dynamic response capability of longitudinal velocity planning, ensuring that vehicles can quickly adjust their velocity planning even in complex and changing traffic environments to avoid collisions and enhance the robustness and safety of the system. In summary, by adopting the embodiments of this application and using a dynamic update method, the system can focus on addressing the most pressing collision risks, avoiding the need to consider all obstacles equally, reducing computational load and processing time, and improving the efficiency of longitudinal velocity generation, thus achieving the technical effect of improving longitudinal velocity generation efficiency.

[0045] As an optional approach, based on the current driving state of the vehicle and its predicted driving state at the first timestamp, a reference longitudinal velocity trajectory is generated, including: Based on the vehicle's driving state and predicted driving state at the current initial timestamp, a first sub-longitudinal velocity trajectory is generated. The first predicted driving state at the first timestamp is calculated when the vehicle is traveling along the first sub-longitudinal velocity trajectory. Based on the first predicted driving state and the second predicted driving state at the third timestamp, a second sub-longitudinal velocity trajectory is generated, where the third timestamp is the timestamp corresponding to the endpoint of the vehicle's longitudinal velocity trajectory. The longitudinal velocity trajectory composed of the first and second sub-longitudinal velocity trajectories is determined as the reference longitudinal velocity trajectory. A target value is added to the number of times the first trajectory is generated to match the reference obstacle object, updating the number of times the first trajectory is generated to a first value, where the number of times the first trajectory is generated represents the number of velocity trajectories generated to avoid the reference obstacle object.

[0046] Optionally, the aforementioned initial timestamp may, but is not limited to, be used to represent the current point in time when the algorithm begins to execute the planning. Starting from this point in time, the planning system analyzes the driving status of the vehicle object and predicts future obstacles.

[0047] Furthermore, the aforementioned predicted driving state may, but is not limited to, the ideal driving state that the vehicle should achieve at a specific point in time (such as the first timestamp) in order to avoid the obstacle, including position, speed, acceleration, etc., to ensure that the vehicle maintains a safe distance from the obstacle.

[0048] Optionally, the aforementioned first sub-longitudinal velocity trajectory may, but is not limited to, represent the longitudinal velocity change path planned for the vehicle object to avoid the reference obstacle object from the initial timestamp to the first timestamp, and the trajectory takes into account the immediate obstacle avoidance requirements.

[0049] It should be noted that the aforementioned first predicted driving state may include, but is not limited to, the state that the vehicle object is expected to reach at the first timestamp based on the first sub-longitudinal velocity trajectory, including longitudinal position, speed and acceleration, in order to evaluate the effectiveness of the obstacle avoidance strategy.

[0050] Furthermore, the aforementioned second sub-longitudinal velocity trajectory may, but is not limited to, refer to the longitudinal velocity change path planned for the vehicle object between the first time stamp and the third time stamp (the trajectory endpoint).

[0051] Optionally, the aforementioned third timestamp may, but is not limited to, be used to represent the end time of the longitudinal velocity trajectory of the vehicle object, that is, the time when the planned lane change operation is completed or the next decision time point is reached.

[0052] It should be noted that the aforementioned reference longitudinal velocity trajectory may include, but is not limited to, a complete longitudinal velocity planning path composed of the first sub-longitudinal velocity trajectory and the second sub-longitudinal velocity trajectory, which takes into account obstacle avoidance requirements and vehicle motion constraints from the current time to a future point in time.

[0053] Furthermore, the number of times the first trajectory is generated may, but is not limited to, refer to a counter for the number of velocity trajectories generated to avoid the highest priority obstacle object (reference obstacle object), in order to control the depth and breadth of the algorithm's exploration and ensure that the generated trajectories are both safe and diverse.

[0054] Optionally, the target value mentioned above can be used, but is not limited to, to represent the increment used to accumulate to the first trajectory generation count (e.g., it can be 1). By iteratively increasing the target value, more obstacle avoidance speed trajectories can be generated, improving the diversity and robustness of the planning.

[0055] It should be noted that the first value mentioned above may include, but is not limited to, the number of times the updated first trajectory has been generated. It reflects the number of obstacle avoidance speed trajectories that the algorithm has generated when processing reference obstacle objects, which helps the algorithm evaluate whether further exploration or optimization of existing trajectories is needed.

[0056] By employing the embodiments of this application, a first sub-longitudinal velocity trajectory and a second sub-longitudinal velocity trajectory are generated, and their combination is determined as a reference longitudinal velocity trajectory. This scheme aims to improve the flexibility and accuracy of speed planning. It ensures that the vehicle can quickly generate effective speed adjustment strategies when facing different obstacles, while controlling the diversity of generated trajectories by controlling the number of times the first trajectory is generated, avoiding the limitations of a single solution. This process can optimize the vehicle's speed control during lane changes, making it more adaptable to dynamically changing traffic environments, while reducing the consumption of computational resources and improving planning speed, thereby achieving safer and faster autonomous driving lane-changing operations.

[0057] As an optional approach, before generating the first sub-longitudinal velocity trajectory based on the vehicle's driving state and predicted driving state at the current initial timestamp, the following steps are also included: The driving state is determined by the vehicle's speed, acceleration, longitudinal distance to the first obstacle at the initial timestamp, and the initial timestamp. The first offset is determined from the candidate offset array using the value corresponding to the number of times the first trajectory is generated as an index. A reference longitudinal distance is calculated using the first offset, the longitudinal safety distance between the vehicle and the first obstacle at the first timestamp, and the first predicted distance. The reference longitudinal distance indicates the longitudinal distance that the vehicle needs to maintain to avoid the first obstacle at the first timestamp. The driving speed, preset acceleration, reference longitudinal distance, and the first timestamp that the vehicle needs to achieve to maintain the reference longitudinal distance with the first obstacle at the first timestamp are determined as the predicted driving state.

[0058] Furthermore, the aforementioned candidate offset array can, but is not limited to, a set of pre-defined values ​​used to fine-tune the trajectory during the generation of the longitudinal velocity trajectory, ensuring that the vehicle can avoid obstacles while also meeting the requirements of comfort and lane-changing efficiency. The selection of the offset increases the diversity of the velocity trajectory, helping the algorithm find the optimal solution.

[0059] It should be noted that the aforementioned first offset may include, but is not limited to, a value from the candidate offset array selected based on the number of times the first trajectory is generated. This value is used to adjust the longitudinal distance between the vehicle object and the first obstacle object to ensure that the generated speed trajectory is both safe and reasonable in terms of obstacle avoidance.

[0060] Furthermore, the aforementioned longitudinal safety distance may, but is not limited to, refer to a minimum longitudinal distance threshold set to ensure that the vehicle remains safe from obstacles during lane changes. This threshold will be dynamically adjusted according to factors such as vehicle speed and acceleration to adapt to different driving conditions and safety requirements.

[0061] Optionally, the aforementioned reference longitudinal distance may, but is not limited to, represent an optimized longitudinal distance that the vehicle object needs to maintain at the first time stamp in order to avoid the first obstacle object. It is calculated based on the longitudinal safety distance and the first offset to ensure that the obstacle avoidance strategy of the vehicle object takes into account both safety, efficiency and driving comfort.

[0062] It should be noted that the predicted driving state mentioned above may include, but is not limited to, the target state of the vehicle at the first timestamp (i.e., the time point when it maintains a reference longitudinal distance from the first obstacle), including information such as the driving speed to be achieved, the preset acceleration, the reference longitudinal distance, and the timestamp. This is a key target point when generating the longitudinal velocity trajectory.

[0063] By employing the embodiments of this application, and accurately determining the driving state, including key parameters such as the vehicle's speed, acceleration, and longitudinal distance from obstacles, as well as selecting the offset based on the number of times the first trajectory is generated, it ensures that the generated speed trajectory not only considers obstacle avoidance requirements but also maintains the comfort and safety of vehicle movement. The process of calculating the reference longitudinal distance and predicting the driving state makes the generation of the speed trajectory more accurate, effectively addressing different obstacle avoidance requirements and enhancing adaptability and robustness to complex scenarios.

[0064] As an optional approach, before generating the second sub-longitudinal velocity trajectory based on the first predicted driving state and the second predicted driving state of the vehicle object at the third timestamp, the following steps are also included: Based on the first value, a first target speed is determined from the candidate longitudinal speed set; based on the preset longitudinal distance configured for the trajectory endpoint and the first target speed, a third timestamp corresponding to the trajectory endpoint is calculated; the preset longitudinal distance, the first target speed, the preset acceleration, and the third timestamp are determined as the second predicted driving state.

[0065] Optionally, the aforementioned set of longitudinal speeds may, but is not limited to, represent a set of preset or calculated longitudinal speed options that provide candidate speed targets for the generation of the second sub-longitudinal speed trajectory based on the vehicle's performance parameters, road conditions, and driving strategy. For example, the set of longitudinal speeds may, but is not limited to, include at least one preset target speed and the current reference obstacle's speed.

[0066] Furthermore, the aforementioned first target speed may, but is not limited to, refer to a specific speed value selected from the candidate longitudinal speed set based on a first numerical value. This speed value is used to guide the longitudinal speed planning of the autonomous vehicle during the lane change process, ensuring that the autonomous vehicle can reach a speed that is both safe and meets the requirements of lane change efficiency after avoiding the first obstacle.

[0067] Optionally, the aforementioned trajectory endpoint may, but is not limited to, be used to represent the final time point and position state of the longitudinal velocity trajectory corresponding to the current planning cycle, i.e., the time and position when the vehicle completes the lane change operation or reaches the next decision point.

[0068] It should be noted that the aforementioned preset longitudinal distance may include, but is not limited to, a preset longitudinal target distance based on the vehicle's performance, road conditions, and driving strategy. This ensures that the vehicle maintains a safe and reasonable distance from the target (such as the next obstacle or target lane) at the end of the trajectory.

[0069] Optionally, the aforementioned preset acceleration may, but is not limited to, be used to represent the acceleration or acceleration range that the vehicle object is expected to adopt during the longitudinal velocity trajectory generation stage. It reflects the vehicle's acceleration or deceleration capabilities, as well as the driving style and comfort requirements desired by the driver or system.

[0070] It should be noted that the aforementioned second predicted driving state may include, but is not limited to, the driving state that the vehicle should achieve at the third time stamp in order to reach the preset longitudinal distance and the first target speed, including parameters such as position, speed, and acceleration, to generate the second sub-longitudinal velocity trajectory, ensuring that the vehicle's journey from avoiding the first obstacle to the end of the trajectory is both safe and meets the expected performance indicators.

[0071] As an optional example, the generation process of the above reference longitudinal velocity trajectory can be illustrated by, but is not limited to, the following example: First, it is necessary to determine the time of the collision. (Used to represent the first timestamp mentioned above) The distance adjusted vertically to avoid the first obstacle (used to represent the first obstacle object). (Used to represent the above reference longitudinal distance), its calculation method is shown in formula (3): (3) in, Indicates that the vehicle is causing the vehicle to be in The minimum longitudinal offset that satisfies the longitudinal safety distance threshold between the object and the obstacle (the first obstacle object) at any given time. The magnitude of the longitudinal offset depends primarily on the degree of overlap. (For example, the longitudinal offset can be defined as the degree of overlap). Furthermore, This represents an additional offset (used to represent the first offset mentioned above). In this example, to ensure the diversity and completeness of the trajectories, an additional offset is added to obtain more valid feasible trajectories, facilitating the selection of the optimal velocity trajectory in subsequent trajectory scoring. The offset can be represented by a one-dimensional array (used to represent the candidate offset array mentioned above), such as... Based on the number of speed trajectories generated for the corresponding obstacle, such as generating speed trajectories for the first obstacle in this case, the remainder of the number divided by 4 is used as the index value of the data to select different offsets. In addition, the positive or negative value of the offset depends on whether the obstacle avoidance strategy is to accelerate or decelerate.

[0072] Then based on the collision time Using these points as dividing points, the velocity trajectories from the current time to the collision time and from the collision time to the final time are calculated using polynomial curves. The velocity trajectories are described by parametric equation (4):

[0073] Assuming the velocity trajectory is divided into k segments, there are k low-order polynomial equations, where n takes the value 4. Considering only the first obstacle, the velocity trajectory is divided into two segments. That is, assuming the velocity trajectory is divided into k segments, there are k low-order polynomial equations, where n takes the value 4. In the current example, considering only the first obstacle, the velocity trajectory is divided into two segments.

[0074] Then, assuming the vehicle's longitudinal displacement is 0 and time is 0 at the current moment, the initial state of the first velocity trajectory (used to represent the first sub-longitudinal velocity trajectory mentioned above) is the vehicle's state (used to represent the driving state) at the current moment. (These represent the longitudinal distance between the vehicle and the first obstacle at the initial timestamp, the corresponding speed, acceleration, and initial timestamp, respectively). Furthermore, the final state of the first velocity trajectory segment is the collision time. The vehicle's state (used to represent the predicted driving state mentioned above) is: (These are used to represent the reference longitudinal distance at the first timestamp, the driving speed required for the vehicle to maintain the reference longitudinal distance from the first obstacle, the preset acceleration, and the first timestamp, respectively), where Assuming the first segment of the trajectory uses a uniformly accelerated trajectory to reach... The final velocity, specifically, It should be noted that since t0 has been configured to 0, therefore... The status of the vehicle In this context, t1 is actually the time difference between t1 and t0.

[0075] Then, based on the initial and final states, the parameters of the parametric equations are solved to obtain the first velocity curve.

[0076] Next, the second velocity trajectory (used to represent the aforementioned second sub-longitudinal velocity trajectory) is calculated. Specifically, the initial state of the second velocity trajectory is obtained based on the equation of the first velocity curve. state of time Among them, state of time This is used to represent the first predicted driving state. The final state of the second speed trajectory is... The final state of the second speed trajectory segment is used to represent the aforementioned second predicted driving state. These are used in sequence to represent the preset longitudinal distance, the first target velocity, the preset acceleration, and the third timestamp. The final longitudinal distance is obtained based on the preset lateral trajectory of the vehicle. To account for the diversity of trajectories, the target speed or the current speed of the obstacle is selected based on the parity of the number of speed trajectories generated for the corresponding obstacle (the two can form the above candidate longitudinal speed set). For example, if the number is odd, the target speed is selected, and if it is even, the current speed of the obstacle is selected; or if the number is odd, the current speed of the obstacle is selected, and if it is even, the target speed is selected. Assuming a uniformly accelerated trajectory with the target as... and The required exercise time.

[0077] For example, the speed planning results generated using different strategies can, but are not limited to, such as... Figure 3 As shown, in the coordinate system containing the curve, the horizontal axis represents time, and the vertical axis represents the longitudinal displacement of the vehicle. Figure 3 The example includes four velocity trajectories. Trajectory 301 uses an accelerated generation strategy, where the additional offset is the larger offset in the above array (e.g., 1 or 2), and the final velocity of the second segment is the target velocity. Trajectory 302 uses an accelerated generation strategy, where the additional offset is the smaller offset in the above array (e.g., 0 or 0.2), and the final velocity of the second segment is the target velocity. Trajectory 303 uses an accelerated generation strategy, where the additional offset is the smaller offset in the above array (e.g., 0 or 0.2), and the final velocity of the second segment is the current obstacle's speed. Trajectory 304 uses a deceleration generation strategy.

[0078] It should be noted that the above examples are optional examples for the purpose of explaining the above steps, and there are no limitations on the above method for generating longitudinal velocity trajectories.

[0079] By employing the embodiments of this application, a first target speed is selected from a set of candidate longitudinal speeds, and the timestamp of the trajectory endpoint is calculated based on a preset longitudinal distance and target speed. This ensures that the generated longitudinal speed trajectory not only meets obstacle avoidance requirements but also considers the endpoint conditions of the speed, including a smooth speed transition and accurate control of the vehicle's final position. This target-based trajectory generation method improves the predictability and controllability of speed planning, contributing to a more efficient and comfortable vehicle driving experience.

[0080] As an optional approach, updating the reference longitudinal velocity trajectory until a target longitudinal velocity trajectory that meets the target conditions is obtained includes: If, based on the reference longitudinal velocity trajectory, a second predicted distance is predicted to be less than the predicted safe distance, and the corresponding second timestamp is earlier than the first timestamp for a second obstacle object, a second offset is determined from the candidate offset array using the first value as an index, and a second target velocity is determined from the candidate longitudinal velocity set using the first value. The predicted driving state is adjusted using the second offset to obtain a third predicted driving state. Based on the driving state, the third predicted driving state, and the second target velocity, a first longitudinal velocity trajectory is generated, and this first longitudinal velocity trajectory is determined as the reference longitudinal velocity trajectory matching the reference obstacle object. The target value is accumulated on the first value to obtain a second value, and the number of times the first trajectory was generated is updated to the second value.

[0081] Optionally, the aforementioned updated reference longitudinal velocity trajectory can, but is not limited to, represent the process of correcting or optimizing the generated reference longitudinal velocity trajectory after a potential collision risk to the vehicle is detected from a new obstacle (a second obstacle object). This ensures that the vehicle can continue to avoid obstacles during subsequent driving while meeting the target conditions of longitudinal velocity planning.

[0082] It should be noted that for the specific implementation of collision detection of each obstacle object based on the reference longitudinal velocity trajectory according to the time sequence, please refer to the embodiments and examples given above, and will not be repeated in this embodiment.

[0083] Furthermore, the aforementioned second predicted distance may include, but is not limited to, the predicted relative distance between the vehicle and the second obstacle at the second timestamp when the vehicle is traveling at its current reference longitudinal speed trajectory. If this distance is less than a preset safety threshold, it indicates a collision risk, and the speed trajectory needs to be adjusted.

[0084] Furthermore, the aforementioned predicted safe distance may refer to, but is not limited to, the minimum longitudinal distance threshold preset by the algorithm to ensure safety between the vehicle and surrounding obstacles. This threshold is dynamically adjusted based on vehicle performance, driving speed, and traffic rules, and is a key criterion for evaluating the effectiveness of the obstacle avoidance strategy. Optionally, the generation method of the lateral and longitudinal predicted safe distances of the second obstacle object at the second timestamp is similar to the generation method of the lateral and longitudinal predicted safe distances of the first obstacle object at the first timestamp described above, and will not be repeated in this embodiment.

[0085] Optionally, the second timestamp can be used, but is not limited to, to represent another point in time that appears in the algorithm's prediction, i.e., the point in time when the vehicle object may collide with the second obstacle object. If the second timestamp is earlier than the first timestamp, it means that after the first obstacle object is avoided, the vehicle object faces another, more urgent obstacle avoidance requirement.

[0086] It should be noted that the aforementioned second offset may include, but is not limited to, another offset value in the candidate offset array selected based on the first value, used to adjust the longitudinal distance between the vehicle object and the second obstacle object, ensuring that the generated speed trajectory can both avoid the first obstacle object and effectively cope with the appearance of the second obstacle object.

[0087] Furthermore, the aforementioned second target speed may, but is not limited to, refer to a target speed value selected from the candidate longitudinal speed set based on the first value, used to set the desired speed of the vehicle when avoiding the second obstacle, so as to optimize the speed trajectory and meet safety and comfort requirements.

[0088] Optionally, the aforementioned third predicted driving state may be used, but is not limited to, to represent the driving state that the vehicle object reaches or needs to maintain at the second timestamp under the adjusted speed trajectory, including position, speed, acceleration, etc., to ensure that the vehicle object can avoid the second obstacle object while following established driving rules and safety standards.

[0089] It should be noted that the aforementioned first longitudinal velocity trajectory may include, but is not limited to, a new longitudinal velocity change path generated for the vehicle object after taking into account the third predicted driving state and the second target speed. This trajectory will serve as a new reference longitudinal velocity trajectory for matching with the reference obstacle object, ensuring that the vehicle avoids all obstacles during lane-changing operations.

[0090] It should be noted that the second value mentioned above may include, but is not limited to, the result of the sum of the first value and the target value. This value represents the number of times the first trajectory was generated after the update, reflecting the cumulative number of speed trajectories generated to avoid the most dangerous obstacle (reference obstacle object).

[0091] Optionally, the above-mentioned adjustment of the predicted driving state using the second offset to obtain the third predicted driving state can be used, but is not limited to, for indication, adjusting the longitudinal displacement of the vehicle in the third predicted driving state based on the second offset, which is the above-mentioned reference longitudinal distance.

[0092] Furthermore, the first longitudinal velocity trajectory generated based on the driving state, the third predicted driving state, and the second target speed can be used, but is not limited to, for indication. In accordance with the generation method of the reference longitudinal velocity trajectory mentioned above, based on the update of the final state of the first trajectory segment, a new longitudinal velocity trajectory is regenerated with the first timestamp as the dividing point and the second target speed as the final state of the second trajectory segment. The specific generation process will not be described in detail in this embodiment.

[0093] By incorporating a second offset and a second target velocity into the embodiments of this application, a new velocity trajectory can be generated, avoiding collisions with new obstacles while maintaining driving continuity and comfort. This process not only enhances the responsiveness to dynamic environments but also increases the diversity of velocity planning strategies by accumulating the number of times the first trajectory is generated, thereby improving the flexibility and safety of the entire planning process.

[0094] As an optional approach, updating the reference longitudinal velocity trajectory until a target longitudinal velocity trajectory that meets the target conditions is obtained also includes: If, based on the reference longitudinal velocity trajectory, the predicted second obstacle is less than the predicted safe distance and the second timestamp is later than the first timestamp, the second obstacle is identified as a candidate reference obstacle. A third sub-longitudinal velocity trajectory is generated based on the predicted driving state and the fourth predicted driving state of the vehicle at the second timestamp, where the fourth predicted driving state is the driving state the vehicle needs to reach at the second timestamp to avoid the second obstacle. A fifth predicted driving state is calculated at the second timestamp when the vehicle is traveling according to the third sub-longitudinal velocity trajectory, and a fourth sub-longitudinal velocity trajectory is generated based on the fifth predicted driving state and the sixth predicted driving state of the vehicle at the fourth timestamp, where the fourth timestamp is the timestamp corresponding to the endpoint of the vehicle's longitudinal velocity trajectory. The longitudinal velocity trajectory composed of the first, third, and fourth sub-longitudinal velocity trajectories is identified as a candidate reference longitudinal velocity trajectory matching the candidate reference obstacle, and a target value is accumulated for the number of times the second trajectory matching the candidate reference obstacle is generated.

[0095] It should be noted that in this embodiment, after generating a reference longitudinal velocity trajectory multiple times with the first obstacle object as the reference obstacle object, if the number of times the first trajectory is generated reaches a first predetermined value, it can be determined that the current scene is not suitable for lane changing, and then the vehicle object is controlled to give up the lane changing operation.

[0096] Optionally, the aforementioned candidate reference obstacle object may, but is not limited to, be used to represent another obstacle detected besides the current reference obstacle object that may affect the lane-changing safety of the vehicle object, which may collide with the vehicle object at the second timestamp and needs to be evaluated and handled.

[0097] It should be noted that the aforementioned fourth predicted driving state may include, but is not limited to, the position, velocity, and acceleration that the vehicle must achieve in order to avoid the second obstacle at the second timestamp. It is the target state for generating the third sub-longitudinal velocity trajectory, ensuring that the vehicle can avoid the second obstacle.

[0098] Furthermore, the aforementioned third longitudinal velocity trajectory may, but is not limited to, refer to the longitudinal velocity change curve generated based on the predicted driving state and the fourth predicted driving state of the vehicle at the first timestamp. This trajectory ensures that the vehicle can not only avoid the first obstacle object during lane changing, but also safely bypass the second obstacle object.

[0099] Optionally, the aforementioned fifth predicted driving state may, but is not limited to, be used to represent the new driving state that the vehicle is expected to reach at the second timestamp when it is traveling along the third sub-longitudinal velocity trajectory. This is a key point for evaluating the effectiveness of the third sub-longitudinal velocity trajectory and ensuring that the vehicle can safely avoid the second obstacle.

[0100] It should be noted that the aforementioned sixth predicted driving state may include, but is not limited to, the position, velocity, and acceleration state that the vehicle must possess at the fourth time stamp (i.e., the end of the trajectory) in order to reach the final longitudinal target position. It is calculated based on the fifth predicted driving state and the preset driving target.

[0101] Furthermore, the aforementioned fourth longitudinal velocity trajectory may, but is not limited to, refer to the longitudinal velocity change path from the second timestamp to the trajectory endpoint generated based on the fifth and sixth predicted driving states. This trajectory ensures that the vehicle can smoothly and efficiently reach the final longitudinal target position after completing the avoidance of the second obstacle.

[0102] Optionally, the number of times the second trajectory is generated can be used, but is not limited to, to represent the number of times the algorithm generates longitudinal velocity trajectories to avoid the candidate reference obstacle object with the highest collision risk and the most need to be processed, excluding the reference obstacle object. By accumulating this value, the depth of trajectory generation can be tracked and controlled to ensure that the planning is both diverse and not too complex, thus affecting real-time performance.

[0103] It should be noted that the second value mentioned above may include, but is not limited to, the number of times the updated second trajectory has been generated. It reflects the number of longitudinal velocity trajectories that have been generated to avoid the second obstacle object, which helps the algorithm manage and control the range of trajectory generation and ensure the efficiency and effectiveness of the algorithm.

[0104] In summary, this embodiment can, but is not limited to, determine the priority of obstacles (i.e., the highest priority reference obstacle and the second highest priority candidate reference obstacle) based on the time when a collision risk exists, and obtain a safe trajectory through iterative collision detection, continuously approaching the optimal speed planning result. This method can quickly determine the target vehicle information for lane changing priority while ensuring trajectory safety.

[0105] Specifically, in this embodiment, the newly generated speed planning curve (e.g., the reference longitudinal speed trajectory generated based on the first obstacle object) is re-calibrated with the obstacle. If a new risky obstacle is detected, different strategies are considered based on the collision time.

[0106] In speed planning, to avoid colliding with other obstacles earlier due to avoiding the first obstacle to collide with, the following steps need to be performed: If the collision time of the new obstacle (e.g., the second obstacle above) If the collision time is less than that of the first obstacle, the additional offset of the collision time and the final state velocity are adjusted according to the number of velocity trajectories generated by the first obstacle, and the trajectory is regenerated (to represent the first longitudinal velocity trajectory mentioned above).

[0107] If the collision time of the new obstacle If the collision time is greater than that of the first obstacle, then the obstacle is designated as the second obstacle (used to represent the aforementioned candidate reference obstacle object).

[0108] At this point, considering both the first and second obstacles simultaneously, the velocity trajectory is divided into three segments. The first segment's velocity curve (used to represent the first sub-longitudinal velocity trajectory) represents the time from the current moment to the collision time with the first obstacle. The original trajectory for avoiding the first obstacle is retained; the second velocity curve (used to represent the longitudinal velocity trajectory of the third sub-object) represents the collision time. Time to the second obstacle collision The third velocity curve (used to represent the longitudinal velocity trajectory of the fourth sub-vehicle) represents the collision time. By the final time It should be noted that the generation method of the new three-segment curves mentioned above is similar to the generation process of the reference longitudinal velocity trajectory described above, and will not be repeated in this embodiment.

[0109] By employing the embodiments of this application, a continuous obstacle avoidance strategy can be formulated for obstacles appearing at different points in time by generating a third sub-longitudinal velocity trajectory and a fourth sub-longitudinal velocity trajectory. This method ensures that the vehicle can continuously avoid multiple obstacles in complex traffic flow while maintaining the rationality and smoothness of the velocity trajectory. By combining multiple sub-trajectories into a complete velocity trajectory and accumulating the number of times the second trajectory is generated, the obstacle avoidance capability and computational efficiency in multi-object scenarios are improved.

[0110] As an optional approach, after determining the longitudinal velocity trajectory composed of the first sub-longitudinal velocity trajectory, the third sub-longitudinal velocity trajectory, and the fourth sub-longitudinal velocity trajectory as the candidate reference longitudinal velocity trajectory matching the candidate reference obstacle object, the following is also included: If a third obstacle object is predicted based on the candidate reference longitudinal velocity trajectory and the predicted distance is less than the predicted safe distance, and the number of times the second trajectory is generated reaches a first predetermined value, the number of times the second trajectory is generated is configured to zero, and the second obstacle object is determined as the reference obstacle object; based on the driving state and the driving state that the vehicle object needs to reach at the second timestamp in order to avoid the second obstacle object, a second longitudinal velocity trajectory is generated; the second longitudinal velocity trajectory is updated as the reference longitudinal velocity trajectory; the target value is accumulated on the first value to obtain the second value, and the number of times the first trajectory is generated is updated to the second value.

[0111] Optionally, the first predetermined value may, but is not limited to, be used to represent a preset threshold value, which is used to generate the number or quantity of velocity trajectories, avoiding unnecessary calculation loops and ensuring the efficiency of the algorithm and resource utilization.

[0112] It should be noted that the aforementioned third predicted distance may include, but is not limited to, the predicted relative distance between the vehicle and the third obstacle at a specific time point (third timestamp) when the vehicle is traveling along the candidate reference longitudinal velocity trajectory. If this distance is less than the safety threshold, it means that the third obstacle may pose a new collision risk, and the velocity trajectory needs to be further adjusted.

[0113] Furthermore, the aforementioned third obstacle may, but is not limited to, other obstacles that may pose a danger to the driving safety of the vehicle, in addition to the first and second obstacles, are detected by collision detection during the vehicle's lane-changing process.

[0114] Optionally, the driving state required to avoid the second obstacle object can be, but is not limited to, representing the target driving state that the vehicle needs to achieve at the second timestamp in order to ensure a safe distance from the second obstacle object, including position, speed, and acceleration, and is the direct basis for generating the second longitudinal velocity trajectory.

[0115] It should be noted that the aforementioned second longitudinal velocity trajectory may include, but is not limited to, a new velocity planning path generated to avoid the second obstacle object. This path takes into account the target driving state of the vehicle at the second timestamp, ensuring that the vehicle can safely and efficiently avoid the second obstacle object during lane changing.

[0116] It should be further noted that the above-mentioned collision detection and new longitudinal velocity trajectory generation process can be referred to the steps for generating the reference longitudinal velocity trajectory above, and will not be repeated in this embodiment.

[0117] In summary, in this embodiment, if the new velocity planning trajectory (used to represent the above-mentioned candidate reference longitudinal velocity trajectory) detects a collision risk with other obstacles, the additional offset of the collision time and the final state velocity are adjusted according to the number of velocity trajectories generated by the second obstacle, and the trajectory is regenerated.

[0118] If the number of speed trajectories generated by the second obstacle (used to represent the number of times the second trajectory is generated) is greater than the preset threshold for the number of trajectories (used to represent the first predetermined value), it is considered that, based on retaining the speed curve of the first obstacle avoidance, it is no longer possible to adjust and generate more speed trajectories that meet the target. Therefore, the second obstacle is replaced with the first obstacle, and the speed trajectory from the current time to time t1 is re-optimized.

[0119] In other words, in this embodiment, based on the collision time of the first obstacle (i.e., the reference obstacle object) and the second obstacle (i.e., the candidate reference obstacle object), the priority of the obstacle is determined, and the speed planning trajectory is constructed in segments. Therefore, the output trajectory has sufficient robustness to cope with complex scenarios. Moreover, by using obstacle iteration verification and priority obstacle replacement, diverse speed trajectories that meet the requirements can be generated while ensuring the safety of the trajectory.

[0120] The preset threshold for the number of trajectories is determined by comprehensively considering the CPU usage, memory consumption, and latency of the planning algorithm while ensuring the diversity of trajectories. In this embodiment, however, no limitation is imposed on this threshold.

[0121] By employing the embodiments of this application, the updating and reconfiguration of the number of times the second trajectory is generated is controlled by setting a first predetermined value. This allows for dynamic adjustment of the obstacle avoidance strategy based on changes in obstacles, ensuring that the vehicle can drive safely and efficiently during continuous obstacle avoidance. It enhances the flexibility and robustness of longitudinal velocity generation, enabling the vehicle to better handle the uncertainties of the traffic environment, while also improving the efficiency of longitudinal velocity generation, providing the vehicle with a better speed planning solution.

[0122] As an alternative approach, the longitudinal velocity trajectory generation method also includes: After updating the reference longitudinal velocity trajectory and obtaining a first candidate longitudinal velocity trajectory that meets the candidate conditions, the first candidate longitudinal velocity trajectory is added to the candidate longitudinal velocity trajectory set. The candidate conditions are used to indicate that when traveling along the candidate longitudinal velocity trajectory within the target time period, the predicted distance to any obstacle object at any timestamp is greater than the predicted safe distance. A third offset is determined from the candidate offset array, and a third target velocity is determined from the candidate longitudinal velocity set. Based on the third offset and the third target velocity, the first candidate longitudinal velocity trajectory is updated to a second candidate longitudinal velocity trajectory. If the second candidate longitudinal velocity trajectory meets the candidate conditions, the second candidate longitudinal velocity trajectory is added to the candidate longitudinal velocity trajectory set.

[0123] Optionally, the aforementioned set of candidate longitudinal velocity trajectories may, but is not limited to, represent a list or data structure that stores all longitudinal velocity trajectories that satisfy the candidate conditions. These trajectories are considered potential optimal solutions and await further evaluation and selection.

[0124] It should be noted that the above candidate conditions may include, but are not limited to, a series of preset performance indicators and constraints. Any generated longitudinal velocity trajectory can only be considered a candidate trajectory if it meets these conditions, especially if it maintains a sufficient safe distance from any obstacle throughout the entire target time period.

[0125] Optionally, the aforementioned third offset may, but is not limited to, represent another offset selected based on the number of trajectory generation times when generating the second candidate longitudinal velocity trajectory. This offset is used to further adjust the longitudinal distance between the vehicle and the obstacle to test the performance and safety of the velocity trajectory under different offsets.

[0126] It should be noted that the aforementioned third target speed may include, but is not limited to, another speed value selected from the candidate longitudinal speed set based on the number of trajectory generation times. It is used as a constraint and optimization objective when generating the second candidate longitudinal speed trajectory, aiming to find multiple solutions that can avoid obstacles and meet performance requirements.

[0127] Optionally, the target time period mentioned above may be used, but is not limited to, to represent the effective time window for the algorithm to perform longitudinal speed planning. It typically covers the time period from the current point in time to the completion of the lane change operation or the arrival of the next decision point, and is used to evaluate the applicability and safety of all longitudinal speed trajectories.

[0128] By employing the embodiments of this application, more speed adjustment schemes can be explored by adding and updating candidate longitudinal speed trajectories, thereby further optimizing driving comfort and efficiency while meeting obstacle avoidance requirements. By accumulating the third offset and the third target speed, and checking the candidate conditions of the generated trajectory, it is ensured that speed planning not only considers safety but also takes into account driving experience and resource utilization, ultimately achieving safe, efficient, and comfortable vehicle lane-changing operations.

[0129] As an alternative approach, the longitudinal velocity trajectory generation method also includes: When updating the reference longitudinal velocity trajectory and obtaining a set of candidate longitudinal velocity trajectories that meet predetermined conditions, the lateral acceleration trajectory and lateral acceleration trajectory of the vehicle object are calculated when controlling the vehicle object based on each candidate longitudinal velocity trajectory in the candidate longitudinal velocity trajectory set. The predetermined conditions are used to indicate that the number of candidate longitudinal velocity trajectories included in the candidate longitudinal velocity trajectory set reaches a predetermined number, or the number of times the first trajectory is generated reaches a second predetermined value, or the number of times the second trajectory is generated reaches a second predetermined value. Based on the lateral acceleration trajectory and lateral jerk trajectory matched with each candidate longitudinal velocity trajectory, obtain the trajectory cost parameters corresponding to each candidate longitudinal velocity trajectory. Based on the trajectory cost parameters corresponding to each candidate longitudinal velocity trajectory, the target longitudinal velocity trajectory that meets the target conditions is determined from the set of candidate longitudinal velocity trajectories.

[0130] Optionally, the aforementioned predetermined conditions may, but are not limited to, represent a set of judgment criteria or thresholds used to determine when to stop generating new longitudinal velocity trajectories and instead proceed to the evaluation phase of the candidate longitudinal velocity trajectory set. These conditions may include the number of trajectories in the candidate longitudinal velocity trajectory set reaching a predetermined number, the number of times the first trajectory is generated reaching a second predetermined value, or the number of times the second trajectory is generated reaching a second predetermined value, ensuring that a sufficient number and diversity of obstacle avoidance velocity trajectories are found within limited time and computational resources, while also preventing infinite loops that could affect the real-time performance and efficiency of the overall planning.

[0131] It should be noted that the aforementioned lateral acceleration trajectory may include, but is not limited to, the acceleration change curve experienced by the vehicle in the lateral direction during lane changing. It is necessary to calculate the impact of each candidate longitudinal velocity trajectory on the vehicle's lateral motion to ensure that the vehicle maintains its lateral dynamics within a safe and comfortable range while executing longitudinal velocity planning.

[0132] Furthermore, the aforementioned lateral acceleration trajectory may, but is not limited to, refer to the curve of change in jerk (i.e., the rate of change of acceleration) experienced by the vehicle along the lateral direction. Similar to the lateral acceleration trajectory, this part of the calculation is also to evaluate the impact of longitudinal velocity planning on the lateral stability of the vehicle, ensuring that the vehicle can smoothly transition when changing lanes and avoid poor passenger experience.

[0133] Optionally, the aforementioned trajectory cost parameters may, but are not limited to, represent a set of quantified metrics used to evaluate the overall performance of each trajectory in the candidate longitudinal velocity trajectory set, including but not limited to safety, comfort, and efficiency. These parameters are calculated based on the lateral acceleration trajectory and lateral jerk trajectory corresponding to each trajectory and are used as the basis for subsequent trajectory selection.

[0134] It should be noted that the aforementioned second predetermined value may include, but is not limited to, a preset upper limit value, used to control the number of times the first trajectory is generated or the number of times the second trajectory is generated, to prevent the algorithm from over-computing, and to ensure that an effective obstacle avoidance speed trajectory can still be generated when resources are limited.

[0135] In other words, in this embodiment, the efficiency of obtaining the optimal speed trajectory is improved by using hard constraint filtering and cost scoring to sort the feasible speed trajectory group. Considering that when the speed planning trajectory detects that there is no risk of collision with any obstacle, the trajectory is recorded in the feasible speed trajectory group. When the number of trajectories in the feasible speed trajectory group is greater than the set threshold (used to represent the above-mentioned second predetermined value) or when the number of speed trajectories generated by only considering the first obstacle is greater than the preset threshold number of trajectories, the feasible speed trajectory group is subjected to hard constraint filtering and cost scoring (used to represent the above-mentioned cost parameters) to sort, thereby outputting the speed planning trajectory with the lowest cost. Among them, the hard constraint provides the maximum safe range of speed, acceleration, and jerk in the speed trajectory according to the motion constraint of the target vehicle, performs preliminary filtering on the speed planning trajectory, reduces the number of trajectories for subsequent cost calculation, and improves calculation efficiency. The calculation method of the cost of the speed trajectory can refer to formula (5):

[0136] Specifically, the velocity planning trajectory is sampled and calculated at 0.1s intervals. This represents the cost of lateral acceleration. This represents the cost of lateral jerk. and The weighting of desired performance can be adjusted by the driving style selected by the driver.

[0137] By employing the embodiments of this application, and through calculating the lateral acceleration trajectory and lateral jerk trajectory, the scheme can comprehensively evaluate the impact of the speed trajectory on the vehicle's lateral dynamics, ensuring that speed planning is not only longitudinally safe but also laterally comfortable. Based on the acquisition and sorting of trajectory cost parameters, the optimal trajectory can be quickly identified from the candidate trajectory set, improving the efficiency and accuracy of the planning process. The setting of predetermined conditions ensures operational efficiency and rational utilization of computing resources in complex scenarios, thereby achieving high performance and high adaptability of the entire speed planning system.

[0138] As an optional example, the above longitudinal velocity trajectory generation method can be illustrated using examples, but not limited to the following: The vehicle is equipped with autonomous driving-related sensors such as LiDAR, millimeter-wave radar, cameras, positioning systems, and inertial measurement units, which can provide the vehicle's positioning, motion, and perception information. The prediction module and the lateral planning module then obtain the predicted trajectory of the obstacle in the next 4 seconds, as well as the lateral trajectory of the vehicle.

[0139] Based on the time sequence, obstacles posing a collision risk to the vehicle are identified according to lateral and longitudinal safety distance thresholds, and their priority is determined based on the collision time. The baseline value for longitudinal safety distance is 1.5, and the baseline value for lateral safety distance is 0.4. Deviation values ​​are calculated based on the current time and the relative speed in both directions. The obstacle that is detected first and has the highest priority is designated as the first obstacle. The obstacle that is the first to collide after avoiding the first obstacle is designated as the second obstacle and has the next highest priority.

[0140] Then, the collision time and location are recorded to determine the required offset for obstacle avoidance. Using the collision time as a dividing point, the piecewise fourth-order polynomial velocity curve equation is solved based on the start and end states to obtain the velocity planning trajectory. The offset is selected from the array [0, 0.2, 1, 2] based on the number of avoidance trajectories generated for the corresponding obstacle, modulo 4. The final velocity at the final point is selected based on the parity of the number of avoidance trajectories generated for the corresponding obstacle; if odd, the target velocity is selected; if even, the obstacle velocity is selected.

[0141] The speed planning trajectory obtained in each round is processed through multiple obstacle collision checks and obstacle priority changes to output diverse speed trajectories that meet the avoidance conditions. Then, the cost of each speed trajectory is calculated, scored, and ranked to output the final optimal speed planning trajectory. For example, when the number of avoidance trajectories generated by the second obstacle is greater than 8, the first obstacle is replaced with the second obstacle, and the trajectory from the current time to the original collision time of the first obstacle is recalculated. When the number of avoidance trajectories generated by the first obstacle is greater than 8, the trajectory generation process is terminated, and the existing speed trajectory group is subjected to hard constraint filtering and cost scoring, and the speed trajectory with the lowest cost is output.

[0142] This example improves computational efficiency while ensuring trajectory safety by setting obstacle priorities and using iterative collision detection. Furthermore, by alternating obstacle priorities and employing different longitudinal offsets, it ensures that the output velocity trajectory does not get trapped in local optima, thus providing a more diverse and abundant range of feasible velocity trajectory solutions for subsequent evaluation and scoring.

[0143] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0144] According to another aspect of the embodiments of this application, a longitudinal velocity trajectory generation apparatus for implementing the above-described longitudinal velocity trajectory generation method is also provided. For example... Figure 7 As shown, the device includes: a first determining unit 702, used to determine obstacle objects included in the driving environment of the vehicle object; a second determining unit 704, used to determine a first obstacle object whose corresponding first predicted distance is less than the predicted safe distance and whose corresponding first timestamp is earlier than the first reference timestamp as a reference obstacle object, wherein the first predicted distance is the predicted distance between the vehicle object and the first obstacle object obtained at the first timestamp; and a generating unit 706, used to generate a reference longitudinal velocity trajectory based on the current driving state of the vehicle object and the predicted driving state of the vehicle object at the first timestamp, wherein the predicted driving state is the trajectory of the vehicle object at the first timestamp to avoid obstacles. The required driving state for the first obstacle object is such that when the vehicle object is driving according to the reference longitudinal speed trajectory at the first time point, the predicted distance between it and the first obstacle object will be greater than the predicted safe distance; the update unit 708 is used to update the reference longitudinal speed trajectory when the second predicted distance based on the reference longitudinal speed trajectory is less than the predicted safe distance and the second time point of the corresponding obstacle object is earlier than the second reference time point, until the target longitudinal speed trajectory that meets the target conditions is obtained. Among them, when the vehicle object is driving according to the target longitudinal speed trajectory within the target time period, the predicted distance between it and any obstacle object at any time point is greater than the predicted safe distance.

[0145] Optionally, in this embodiment, the generation unit includes: a first generation module, used to generate a first sub-longitudinal speed trajectory based on the driving state and predicted driving state of the vehicle object at the current corresponding initial timestamp; a second generation module, used to calculate the first predicted driving state corresponding to the first timestamp when the vehicle object drives according to the first sub-longitudinal speed trajectory, and generate a second sub-longitudinal speed trajectory based on the first predicted driving state and the second predicted driving state corresponding to the vehicle object at the third timestamp, wherein the third timestamp is the timestamp corresponding to the trajectory endpoint of the longitudinal speed trajectory of the vehicle object; a first determination module, used to determine the longitudinal speed trajectory composed of the first sub-longitudinal speed trajectory and the second sub-longitudinal speed trajectory as a reference longitudinal speed trajectory; and a first update module, used to accumulate a target value for the number of times the first trajectory is generated to match the reference obstacle object, so as to update the number of times the first trajectory is generated to a first value, wherein the number of times the first trajectory is generated is used to characterize the number of speed trajectories generated to avoid the reference obstacle object.

[0146] Optionally, in this embodiment, the above-mentioned device further includes: a third determining unit, used to determine the driving speed, driving acceleration, longitudinal distance between the vehicle object and the first obstacle object at the initial timestamp, and the initial timestamp as the driving state; a fourth determining unit, used to determine the first offset from the candidate offset array using the value corresponding to the number of times the first trajectory is generated as an index; a first calculation unit, used to calculate a reference longitudinal distance using the first offset, the longitudinal safety distance between the vehicle object and the first obstacle object at the first timestamp, and the first predicted distance, wherein the reference longitudinal distance is used to indicate the longitudinal distance that the vehicle object needs to maintain to avoid the first obstacle object and the first obstacle at the first timestamp; and a fifth determining unit, used to determine the driving speed, preset acceleration, reference longitudinal distance, and the first timestamp that the vehicle object needs to achieve to maintain the reference longitudinal distance with the first obstacle object at the first timestamp as the predicted driving state.

[0147] Optionally, in this embodiment, the above-mentioned device further includes: a sixth determining unit, configured to determine a first target speed from a set of candidate longitudinal speeds based on a first value; a second calculating unit, configured to calculate a third timestamp corresponding to the trajectory endpoint based on a preset longitudinal distance configured for the trajectory endpoint and the first target speed; and a seventh determining unit, configured to determine the preset longitudinal distance, the first target speed, the preset acceleration, and the third timestamp as a second predicted driving state.

[0148] Optionally, in this embodiment, the above-mentioned updating unit includes: a second determining module, used to determine a second offset from a candidate offset array using a first value as an index, and to determine a second target speed from a candidate longitudinal speed set using the first value, when a second obstacle object is predicted based on a reference longitudinal speed trajectory and the corresponding second predicted distance is less than the predicted safe distance and the corresponding second timestamp is earlier than the first timestamp, based on the second obstacle object; a first adjusting module, used to adjust the predicted driving state using the second offset to obtain a third predicted driving state; a third generating module, used to generate a first longitudinal speed trajectory based on the driving state, the third predicted driving state, and the second target speed, and to determine the first longitudinal speed trajectory as a reference longitudinal speed trajectory matching the reference obstacle object; and a second updating module, used to accumulate the target value on the first value to obtain a second value, and to update the number of times the first trajectory was generated to the second value.

[0149] Optionally, in this embodiment, the above-mentioned device further includes: an eighth determining unit, configured to determine the second obstacle object as a candidate reference obstacle object when the second predicted distance is less than the predicted safe distance and the second timestamp is later than the first timestamp, based on the reference longitudinal velocity trajectory; a first generating unit, configured to generate a third sub-longitudinal velocity trajectory based on the predicted driving state and the fourth predicted driving state corresponding to the vehicle object at the second timestamp, wherein the fourth predicted driving state is the driving state that the vehicle object needs to achieve at the second timestamp in order to avoid the second obstacle object; and a second generating unit, configured to calculate the vehicle object according to... In the case of driving along the third sub-longitudinal velocity trajectory, the fifth predicted driving state corresponding to the second timestamp is used, and based on the fifth predicted driving state and the sixth predicted driving state corresponding to the vehicle object at the fourth timestamp, a fourth sub-longitudinal velocity trajectory is generated, wherein the fourth timestamp is the timestamp corresponding to the trajectory endpoint of the vehicle object's longitudinal velocity trajectory; the ninth determining unit is used to determine the longitudinal velocity trajectory composed of the first sub-longitudinal velocity trajectory, the third sub-longitudinal velocity trajectory, and the fourth sub-longitudinal velocity trajectory as a candidate reference longitudinal velocity trajectory that matches the candidate reference obstacle object, and to accumulate a target value for the number of times the second trajectory matching the candidate reference obstacle object is generated.

[0150] Optionally, in this embodiment, the above-mentioned device further includes: a tenth determining unit, configured to set the second trajectory generation count to zero and determine the second obstacle object as a reference obstacle object when a third obstacle object with a predicted third distance less than the predicted safe distance is predicted based on the candidate reference longitudinal velocity trajectory, and the second trajectory generation count reaches a first predetermined value; a third generating unit, configured to generate a second longitudinal velocity trajectory based on the driving state and the driving state that the vehicle object needs to reach at the second timestamp in order to avoid the second obstacle object; a first updating unit, configured to update the second longitudinal velocity trajectory as the reference longitudinal velocity trajectory; and a second updating unit, configured to accumulate a target value on the first value to obtain a second value and update the first trajectory generation count to the second value.

[0151] Optionally, in this embodiment, the above-mentioned device further includes: an adding unit, configured to add the first candidate longitudinal velocity trajectory to the candidate longitudinal velocity trajectory set when the reference longitudinal velocity trajectory is updated and a first candidate longitudinal velocity trajectory that meets the candidate conditions is obtained, wherein the candidate conditions are used to indicate that when traveling according to the candidate longitudinal velocity trajectory within the target time period, the predicted distance between the vehicle and any obstacle object at any timestamp is greater than the predicted safe distance; determine a third offset from the candidate offset array and determine a third target speed from the candidate longitudinal velocity set; update the first candidate longitudinal velocity trajectory to a second candidate longitudinal velocity trajectory based on the third offset and the third target speed; and add the second candidate longitudinal velocity trajectory to the candidate longitudinal velocity trajectory set when the second candidate longitudinal velocity trajectory meets the candidate conditions.

[0152] Optionally, in this embodiment, the above-mentioned device further includes: a target longitudinal velocity trajectory determination unit, configured to: when updating the reference longitudinal velocity trajectory to obtain a set of candidate longitudinal velocity trajectories that meet predetermined conditions, calculate the lateral acceleration trajectory and lateral jerk trajectory corresponding to the vehicle object when controlling the vehicle object based on each candidate longitudinal velocity trajectory in the set of candidate longitudinal velocity trajectories, wherein the predetermined conditions are used to indicate that the number of candidate longitudinal velocity trajectories included in the set of candidate longitudinal velocity trajectories reaches a predetermined number, or the number of times the first trajectory is generated reaches a second predetermined value, or the number of times the second trajectory is generated reaches a second predetermined value; obtain the trajectory cost parameter corresponding to each candidate longitudinal velocity trajectory based on the lateral acceleration trajectory and lateral jerk trajectory matched with each candidate longitudinal velocity trajectory; and determine the target longitudinal velocity trajectory that meets the target conditions from the set of candidate longitudinal velocity trajectories based on the trajectory cost parameter corresponding to each candidate longitudinal velocity trajectory.

[0153] For specific implementation examples, please refer to the examples shown in the above longitudinal velocity trajectory generation method. This embodiment will not be repeated here.

[0154] According to another aspect of the embodiments of this application, an electronic device for implementing the above-described longitudinal velocity trajectory generation method is also provided. This electronic device can be an in-vehicle terminal device or a server. This embodiment uses an in-vehicle terminal device as an example for illustration. Figure 8 As shown, the electronic device includes a memory 802 and a processor 804. The memory 802 stores a computer program, and the processor 804 is configured to execute the steps in any of the above method embodiments via the computer program.

[0155] Optionally, in this embodiment, the aforementioned electronic device may be located in at least one of a plurality of network devices in a computer network.

[0156] Optionally, in this embodiment, the processor can be configured to execute the relevant steps of the longitudinal velocity trajectory generation method via a computer program.

[0157] Alternatively, as those skilled in the art will understand, Figure 8 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones (such as Android phones, iOS phones, etc.), tablets, PDAs, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 8 This does not limit the structure of the aforementioned electronic devices. For example, the electronic device may also include components that are more... Figure 8 The more or fewer components shown (such as network interfaces, etc.), or having the same Figure 8 The different configurations shown.

[0158] The memory 802 can be used to store software programs and modules, such as the program instructions / modules corresponding to the longitudinal velocity trajectory generation method and apparatus in this embodiment. The processor 804 executes various functional applications and data processing by running the software programs and modules stored in the memory 802, thereby realizing the aforementioned longitudinal velocity trajectory generation method. The memory 802 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 802 may further include memory remotely located relative to the processor 804, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. As an example, such as Figure 8As shown, the memory 802 may include, but is not limited to, the first determining unit 702, the second determining unit 704, the generating unit 706, and the updating unit 708 from the longitudinal velocity trajectory generating device. Furthermore, it may include, but is not limited to, other module units from the longitudinal velocity trajectory generating device, which will not be elaborated upon in this example.

[0159] Optionally, the transmission device 806 described above is used to receive or send data via a network. Specific examples of the network described above may include wired networks and wireless networks. In one example, the transmission device 806 includes a Network Interface Controller (NIC), which can be connected to other network devices and a router via a network cable to communicate with the Internet or a local area network. In another example, the transmission device 806 is a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0160] In addition, the above-mentioned electronic device also includes: a display 808 and a connection bus 810 for connecting the various module components in the above-mentioned electronic device.

[0161] According to one aspect of this application, a computer program product is provided, comprising a computer program / instructions containing program code for performing the methods described above. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit, it performs various functions provided in the embodiments of this application.

[0162] According to one aspect of this application, another computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the methods in various embodiments of this application.

[0163] According to one aspect of this application, a computer-readable storage medium is provided, wherein a processor of a computer device reads computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the above-described method.

[0164] Optionally, in this embodiment, the computer-readable storage medium described above may be configured to store steps for performing the methods described above.

[0165] It should be noted that the data collection and processing described in this application should strictly comply with the requirements of relevant national laws and regulations, obtain the informed consent or separate consent of the personal information subject, and carry out subsequent data use and processing within the scope of laws and regulations and the authorization of the personal information subject.

[0166] Optionally, in the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program with a predetermined function, which works together with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0167] Optionally, in this embodiment, those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0168] If the integrated units in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0169] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0170] In the several embodiments provided in this application, it should be understood that the disclosed client can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection between units or modules, and may be electrical or other forms.

[0171] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0172] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0173] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for generating a longitudinal velocity trajectory, characterized in that, include: Identify the obstacles in the driving environment of the vehicle. The first obstacle object whose corresponding first predicted distance is less than the predicted safe distance and whose corresponding first timestamp is earlier than the first reference timestamp is determined as the reference obstacle object, wherein the first predicted distance is the predicted distance between the vehicle object and the first obstacle object obtained at the first timestamp; Based on the driving state of the vehicle at the current initial timestamp and the predicted driving state of the vehicle at the first timestamp, a first sub-longitudinal velocity trajectory is generated, wherein the predicted driving state is the driving state that the vehicle needs to achieve at the first timestamp in order to avoid the first obstacle object. Calculate the first predicted driving state corresponding to the first timestamp when the vehicle object is traveling according to the first sub-longitudinal speed trajectory, and generate a second sub-longitudinal speed trajectory based on the first predicted driving state and the second predicted driving state corresponding to the vehicle object at the third timestamp, wherein the third timestamp is the timestamp corresponding to the trajectory endpoint of the longitudinal speed trajectory of the vehicle object. The longitudinal velocity trajectory formed by the first sub-longitudinal velocity trajectory and the second sub-longitudinal velocity trajectory is determined as the reference longitudinal velocity trajectory. When the vehicle is traveling according to the reference longitudinal velocity trajectory at the first timestamp, the predicted distance between it and the first obstacle will be greater than the predicted safe distance. An accumulation target value is added to the number of times a first trajectory is generated to match the reference obstacle object, so as to update the number of times the first trajectory is generated to a first value, wherein the number of times the first trajectory is generated is used to characterize the number of speed trajectories generated to avoid the reference obstacle object; If, based on the reference longitudinal velocity trajectory, the predicted second distance is less than the predicted safe distance and the corresponding second timestamp is earlier than the second reference timestamp of the second obstacle object, the reference longitudinal velocity trajectory is updated until a target longitudinal velocity trajectory that meets the target conditions is obtained. Wherein, when the vehicle object travels according to the target longitudinal velocity trajectory within the target time period, the predicted distance between it and any obstacle object at any timestamp is greater than the predicted safe distance.

2. The method according to claim 1, characterized in that, Before generating the first sub-longitudinal velocity trajectory based on the driving state of the vehicle object at the current corresponding initial timestamp and the predicted driving state of the vehicle object at the first timestamp, the method further includes: The driving state is determined by the vehicle's speed, acceleration, longitudinal distance to the first obstacle object at the initial timestamp, and the initial timestamp. Using the value corresponding to the number of times the first trajectory was generated as an index, the first offset is determined from the candidate offset array; Using the first offset, the longitudinal safe distance between the vehicle and the first obstacle at the first timestamp, and the first predicted distance, a reference longitudinal distance is calculated, wherein the reference longitudinal distance is used to indicate the longitudinal distance that the vehicle needs to maintain in order to avoid the first obstacle at the first timestamp and the first obstacle. The predicted driving state is determined by the driving speed, preset acceleration, reference longitudinal distance, and the first timestamp required by the vehicle to maintain the reference longitudinal distance with the first obstacle.

3. The method according to claim 2, characterized in that, Before generating the second sub-longitudinal velocity trajectory based on the first predicted driving state and the second predicted driving state corresponding to the vehicle object at the third timestamp, the method further includes: Based on the first value, a first target velocity is determined from the candidate longitudinal velocity set; Based on the preset longitudinal distance configured for the trajectory endpoint and the first target velocity, the third timestamp corresponding to the trajectory endpoint is calculated; The preset longitudinal distance, the first target speed, the preset acceleration, and the third timestamp are determined as the second predicted driving state.

4. The method according to claim 3, characterized in that, The step of updating the reference longitudinal velocity trajectory until a target longitudinal velocity trajectory that meets the target conditions is obtained includes: If, based on the reference longitudinal velocity trajectory, the predicted second distance is less than the predicted safe distance and the corresponding second timestamp is earlier than the first timestamp for the second obstacle object, the second offset is determined from the candidate offset array using the first value as an index, and the second target velocity is determined from the candidate longitudinal velocity set using the first value. The predicted driving state is adjusted using the second offset to obtain the third predicted driving state; Based on the driving state, the third predicted driving state, and the second target speed, a first longitudinal speed trajectory is generated, and the first longitudinal speed trajectory is determined as a reference longitudinal speed trajectory that matches the reference obstacle object. The target value is added to the first value to obtain the second value, and the number of times the first trajectory is generated is updated to the second value.

5. The method according to claim 3, characterized in that, The step of updating the reference longitudinal velocity trajectory until a target longitudinal velocity trajectory that meets the target conditions is obtained further includes: If, based on the reference longitudinal velocity trajectory, the second obstacle object is predicted to have a second predicted distance that is less than the predicted safe distance and a second timestamp that is later than the first timestamp, then the second obstacle object is identified as a candidate reference obstacle object. Based on the predicted driving state and the fourth predicted driving state of the vehicle at the second timestamp, a third sub-longitudinal velocity trajectory is generated, wherein the fourth predicted driving state is the driving state that the vehicle needs to achieve at the second timestamp in order to avoid the second obstacle object. Calculate the fifth predicted driving state corresponding to the second timestamp when the vehicle object is traveling according to the third sub-longitudinal velocity trajectory, and generate the fourth sub-longitudinal velocity trajectory based on the fifth predicted driving state and the sixth predicted driving state corresponding to the vehicle object at the fourth timestamp, wherein the fourth timestamp is the timestamp corresponding to the trajectory endpoint of the longitudinal velocity trajectory of the vehicle object. The longitudinal velocity trajectory composed of the first sub-longitudinal velocity trajectory, the third sub-longitudinal velocity trajectory, and the fourth sub-longitudinal velocity trajectory is determined as the candidate reference longitudinal velocity trajectory that matches the candidate reference obstacle object, and the number of times the second trajectory that matches the candidate reference obstacle object is generated is increased by a target value.

6. The method according to claim 5, characterized in that, After determining the longitudinal velocity trajectory composed of the first sub-longitudinal velocity trajectory, the third sub-longitudinal velocity trajectory, and the fourth sub-longitudinal velocity trajectory as the candidate reference longitudinal velocity trajectory matching the candidate reference obstacle object, the method further includes: If a third obstacle object with a predicted third distance less than the predicted safe distance is predicted based on the candidate reference longitudinal velocity trajectory, and the number of times the second trajectory is generated reaches a first predetermined value, the number of times the second trajectory is generated is configured to zero, and the second obstacle object is determined as the reference obstacle object. Based on the driving state and the driving state that the vehicle needs to achieve at the second timestamp in order to avoid the second obstacle, a second longitudinal velocity trajectory is generated. Update the second longitudinal velocity trajectory to the reference longitudinal velocity trajectory; The target value is accumulated on the first value to obtain the second value, and the number of times the first trajectory is generated is updated to the second value.

7. The method according to claim 3, characterized in that, The longitudinal velocity trajectory generation method further includes: When the reference longitudinal velocity trajectory is updated and a first candidate longitudinal velocity trajectory that meets the candidate conditions is obtained, the first candidate longitudinal velocity trajectory is added to the candidate longitudinal velocity trajectory set. The candidate conditions are used to indicate that when traveling according to the candidate longitudinal velocity trajectory within the target time period, the predicted distance between the vehicle and any obstacle object at any time stamp is greater than the predicted safe distance. The third offset is determined from the candidate offset array, and the third target velocity is determined from the candidate longitudinal velocity set; Based on the third offset and the third target velocity, the first candidate longitudinal velocity trajectory is updated to the second candidate longitudinal velocity trajectory; If the second candidate longitudinal velocity trajectory meets the candidate conditions, the second candidate longitudinal velocity trajectory is added to the candidate longitudinal velocity trajectory set.

8. The method according to claim 5, characterized in that, The longitudinal velocity trajectory generation method further includes: When updating the reference longitudinal velocity trajectory to obtain a set of candidate longitudinal velocity trajectories that meet predetermined conditions, the lateral acceleration trajectory and lateral acceleration trajectory corresponding to the vehicle object are calculated when controlling the vehicle object based on each of the candidate longitudinal velocity trajectories in the set of candidate longitudinal velocity trajectories. The predetermined conditions are used to indicate that the number of candidate longitudinal velocity trajectories included in the set of candidate longitudinal velocity trajectories reaches a predetermined number, or the number of times the first trajectory is generated reaches a second predetermined value, or the number of times the second trajectory is generated reaches a second predetermined value. Based on the lateral acceleration trajectory and lateral jerk trajectory that match each of the candidate longitudinal velocity trajectories, obtain the trajectory cost parameters corresponding to each of the candidate longitudinal velocity trajectories. Based on the trajectory cost parameters corresponding to each of the candidate longitudinal velocity trajectories, the target longitudinal velocity trajectory that satisfies the target condition is determined from the set of candidate longitudinal velocity trajectories.

9. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the method described in any one of claims 1 to 8 through the computer program.

Images