Trajectory Correction Method, Device, Cloud Control Platform, and Autonomous Driving Vehicle

Abstract

The present disclosure provides a trajectory correction method, device, cloud control platform, and autonomous vehicle, which relate to the field of artificial intelligence, and particularly to the fields of autonomous driving and intelligent transportation technologies. The specific implementation scheme includes: determining a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments, where M is an integer greater than 0; determining a trajectory correction parameter according to the predicted obstacle state and the predicted obstacle trajectory for the target time period; and correcting the predicted obstacle trajectory based on the trajectory correction parameter to obtain a corrected target obstacle trajectory, where the predicted obstacle trajectory is obtained based on obstacle state data for N historical moments, N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence, and particularly to the fields of autonomous driving and intelligent transportation technologies, and can be applied to scenarios such as trajectory correction. Background Art

[0002] In the field of autonomous driving, obstacle trajectory prediction can provide effective decision-making support for driving assistance control. However, in some scenarios, there are phenomena of low data utilization rate and poor prediction accuracy in the trajectory prediction process. Summary of the Invention

[0003] The present disclosure provides a trajectory correction method, device, cloud control platform, and autonomous driving vehicle.

[0004] According to one aspect of the present disclosure, there is provided a trajectory correction method, including: determining a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments, where M is an integer greater than 0; determining a trajectory correction parameter according to the predicted obstacle state and a predicted obstacle trajectory for the target time period; and correcting the predicted obstacle trajectory based on the trajectory correction parameter to obtain a corrected target obstacle trajectory, where the predicted obstacle trajectory is obtained based on obstacle state data of N historical moments, N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0005] According to another aspect of the present disclosure, there is provided a trajectory correction device, including: a first processing module for determining a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments, where M is an integer greater than 0; a second processing module for determining a trajectory correction parameter according to the predicted obstacle state and a predicted obstacle trajectory for the target time period; and a third processing module for correcting the predicted obstacle trajectory based on the trajectory correction parameter to obtain a corrected target obstacle trajectory, where the predicted obstacle trajectory is obtained based on obstacle state data of N historical moments, N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0006] According to another aspect of the present disclosure, there is provided an electronic device, including: at least one processor and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the trajectory correction method of the above aspect.

[0007] According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the trajectory correction method of the above aspect.

[0008] According to another aspect of the present disclosure, there is provided a computer program product including a computer program which, when executed by a processor, implements the trajectory correction method of the above aspect.

[0009] According to another aspect of the present disclosure, there is provided a cloud control platform including the electronic device of the above aspect.

[0010] According to another aspect of the present disclosure, there is provided an autonomous vehicle including the electronic device of the above aspect.

[0011] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0013] Figure 1 Schematically shows the system architecture of the trajectory correction method and apparatus according to an embodiment of the present disclosure;

[0014] Figure 2 Schematically shows the flowchart of the trajectory correction method according to an embodiment of the present disclosure;

[0015] Figure 3 Schematically shows the flowchart of the trajectory correction method according to another embodiment of the present disclosure;

[0016] Figure 4 Schematically shows the schematic diagram of the trajectory correction process according to an embodiment of the present disclosure;

[0017] Figure 5 Schematically shows the block diagram of the trajectory correction apparatus according to an embodiment of the present disclosure;

[0018] Figure 6 Schematically shows the block diagram of the electronic device for executing trajectory correction according to an embodiment of the present disclosure. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The following describes exemplary embodiments of the present disclosure with reference to the accompanying drawings. Various details of the embodiments of the present disclosure are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, descriptions of well-known functions and structures are omitted in the following description for clarity and conciseness.

[0020] The terms used herein are merely for describing specific embodiments and are not intended to limit the present disclosure. The terms "including", "comprising", etc. used herein indicate the presence of the described features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0021] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those of ordinary skill in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0022] In cases where expressions similar to "at least one of A, B, and C, etc." are used, generally, it should be interpreted according to the meaning commonly understood by those of ordinary skill in the art (for example, "a system having at least one of A, B, and C" should include, but is not limited to, a system having only A, only B, only C, having A and B, having A and C, having B and C, and / or having A, B, and C, etc.).

[0023] An embodiment of the present disclosure provides a trajectory correction method. The trajectory correction method includes: determining a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments, where M is an integer greater than 0; determining a trajectory correction parameter according to the predicted obstacle state and the predicted obstacle trajectory for the target time period; and correcting the predicted obstacle trajectory based on the trajectory correction parameter to obtain a corrected target obstacle trajectory. The predicted obstacle trajectory is obtained based on obstacle state data of N historical moments, where N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0024] Figure 1 The system architecture of the trajectory correction method and apparatus according to an embodiment of the present disclosure is schematically shown. It should be noted that Figure 1 The figure shown is only an example of the system architecture to which the embodiments of the present disclosure can be applied, to help those of ordinary skill in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments, or scenarios.

[0025] The system architecture 100 according to this embodiment may include multiple obstacles (as Figure 1 shown, for example, including obstacles 101, 102, and 103), a network 104, and a server 105. The network 104 is used to provide a medium for communication links between the obstacles and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc. The server 105 may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud computing, network services, and middleware services. The obstacles may be other vehicles around the target vehicle, for example, other vehicles within the coverage range when the target vehicle acquires environmental data.

[0026] For any of the obstacles 101, 102, and 103, the server 105 can determine the predicted obstacle state for the target time period associated with the corresponding obstacle according to the obstacle state data based on M historical moments, where M is an integer greater than 0. According to the predicted obstacle state and the predicted obstacle trajectory for the target time period, determine the trajectory correction parameter, and based on the trajectory correction parameter, correct the predicted obstacle trajectory to obtain the corrected target obstacle trajectory. The predicted obstacle trajectory is obtained based on the obstacle state data of N historical moments, where N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0027] In one example, an autonomous vehicle includes an electronic device, and the electronic device includes but is not limited to a vehicle-mounted system. The electronic device can execute the trajectory correction method of the embodiments of the present disclosure. Exemplarily, the vehicle-mounted system of the autonomous vehicle may have a data processing function, and the vehicle-mounted system can perform operations based on the obstacle state data to obtain the corrected target obstacle trajectory.

[0028] In another example, a cloud control platform includes an electronic device, and the electronic device can execute the trajectory correction method of the embodiments of the present disclosure.

[0029] It should be understood that Figure 1 the numbers of obstacles, networks, and servers in

[0030] It should be noted that the trajectory correction method provided by the embodiments of the present disclosure can be executed by the server 105 or the autonomous vehicle. Correspondingly, the trajectory correction device provided by the embodiments of the present disclosure can be disposed in the server 105 or the autonomous vehicle. The trajectory correction method provided by the embodiments of the present disclosure can also be executed by a server or a server cluster different from the server 105 and capable of communicating with the autonomous vehicle and / or the server 105. Correspondingly, the trajectory correction device provided by the embodiments of the present disclosure can also be disposed in a server or a server cluster different from the server 105 and capable of communicating with the autonomous vehicle and / or the server 105.

[0031] It should be noted that in the technical solution of the present disclosure, the processing of collection, storage, use, processing, transmission, provision, and disclosure of the user's personal information involved all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0032] In the technical solution of the present disclosure, before obtaining or collecting the user's personal information, the authorization or consent of the user has been obtained.

[0033] The embodiments of the present disclosure provide a trajectory correction method. The following will be combined with Figure 1 the system architecture, and with reference to Figures 2 to 4 to describe the trajectory correction method according to the exemplary embodiments of the present disclosure. The trajectory correction method of the embodiments of the present disclosure can be executed, for example, by Figure 1 the server 105 shown.

[0034] Figure 2 Schematically shows a flowchart of the trajectory correction method according to an embodiment of the present disclosure.

[0035] As Figure 2 shown, the trajectory correction method 200 of the embodiments of the present disclosure may include, for example, operation S210 to operation S230.

[0036] In operation S210, in response to the obstacle state data based on M historical moments, a predicted obstacle state for the target time period is determined, where M is an integer greater than 0.

[0037] In operation S220, trajectory correction parameters are determined according to the predicted obstacle state and the predicted obstacle trajectory for the target time period.

[0038] In operation S230, based on the trajectory correction parameters, the predicted obstacle trajectory is corrected to obtain a corrected target obstacle trajectory.

[0039] The predicted obstacle trajectory is obtained based on the obstacle state data of N historical moments, where N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0040] The following example illustrates the example process of each operation of the trajectory correction method in this embodiment.

[0041] Exemplarily, in response to the obstacle state data based on M historical moments, a predicted obstacle state for a target time period is determined, where M is an integer greater than 0. The obstacle may be other vehicles around the target vehicle. For example, it may be other vehicles within the range covered when the target vehicle acquires environmental data, and the other vehicles can be considered as dynamic obstacles in the driving environment of the target vehicle.

[0042] The obstacle state data may include, for example, obstacle motion parameters, vehicle control parameters, obstacle position parameters, obstacle attribute parameters, and driving lane parameters, etc. Based on the obstacle state data of M historical moments, a predicted obstacle state for the target time period is determined. The M historical moments may be at least one historical moment adjacent to the target time period. The predicted obstacle state may include, for example, at least one of the following driving parameters: obstacle driving speed, obstacle driving acceleration, and obstacle driving steering angle.

[0043] Exemplarily, the obstacle motion parameters may include content such as obstacle speed parameters, obstacle acceleration parameters, obstacle steering angle parameters, etc. The vehicle control parameters may indicate the control state of the obstacle. For example, they may indicate information such as the braking state, turn signal state, and steering motor torque of the obstacle. The obstacle position parameters may indicate the position coordinates of the obstacle. For example, they may indicate the longitude and latitude coordinates of the obstacle, and the longitude and latitude coordinates may correspond to map elements in a high-precision map. The obstacle attribute parameters may indicate information such as obstacle type, obstacle width, and obstacle length. The driving lane parameters may indicate information such as lane width, lane line slope, and lateral distance between the obstacle and the lane line.

[0044] Based on the predicted obstacle state and the predicted obstacle trajectory for the target time period, a trajectory correction parameter is determined. The predicted obstacle trajectory is obtained based on the obstacle state data of N historical moments, where N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0045] The target time period may include at least one target moment starting from the current moment, and the N historical moments may include at least one historical moment ending at the current moment. The M historical moments may include a partial historical moment closest to the current moment among the N historical moments.

[0046] In an example method, when determining a trajectory correction parameter according to a predicted obstacle state and a predicted obstacle trajectory for a target time period, a reference obstacle position based on at least one target moment within the target time period may be determined according to the predicted obstacle state, and the trajectory correction parameter may be determined according to the reference obstacle position and a predicted obstacle position based on at least one target moment indicated by the predicted obstacle trajectory.

[0047] Based on the trajectory correction parameter, the predicted obstacle trajectory is corrected to obtain a corrected target obstacle trajectory. Exemplarily, according to the trajectory correction parameter, the predicted obstacle position based on at least one target moment is corrected to obtain a corrected target obstacle trajectory. Based on the corrected target obstacle trajectory, a vehicle control instruction is generated, and the vehicle control instruction is sent to a vehicle control terminal so that the vehicle can be controlled to travel based on the vehicle control instruction, which can effectively improve the vehicle control accuracy and effectively ensure the safe driving of the autonomous vehicle.

[0048] In an embodiment of the present disclosure, in response to obstacle state data based on M historical moments, a predicted obstacle state for a target time period is determined, where M is an integer greater than 0. According to the predicted obstacle state and a predicted obstacle trajectory for the target time period, a trajectory correction parameter is determined, and based on the trajectory correction parameter, the predicted obstacle trajectory is corrected to obtain a corrected target obstacle trajectory. The predicted obstacle trajectory is obtained based on obstacle state data of N historical moments, where N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0049] Based on obstacle state data of at least one historical moment adjacent to the target time period, a predicted obstacle state associated with the target time period is determined, and based on the predicted obstacle state and a predetermined predicted obstacle trajectory, a trajectory correction parameter is determined. By improving the utilization rate of obstacle state data, the accuracy of obstacle trajectory prediction can be effectively improved, the accuracy of autonomous driving control can be effectively enhanced, credible decision support can be provided for driving assistance control, and it is beneficial to ensure the safe driving of the autonomous vehicle.

[0050] Figure 3 A flowchart of a trajectory correction method according to another embodiment of the present disclosure is schematically shown.

[0051] As Figure 3 shown, the trajectory correction method 300 of the embodiment of the present disclosure may include operations S210, S310, and S230, for example.

[0052] In operation S210, in response to obstacle state data based on M historical moments, a predicted obstacle state for a target time period is determined, where M is an integer greater than 0.

[0053] In operation S310, based on the predicted obstacle state, a reference obstacle position based on at least one target time in the target time period is determined, and a trajectory correction parameter is determined according to the reference obstacle position and the predicted obstacle position based on at least one target time indicated by the predicted obstacle trajectory.

[0054] In operation S230, the predicted obstacle trajectory is corrected based on the trajectory correction parameter to obtain a corrected target obstacle trajectory.

[0055] The predicted obstacle trajectory is obtained based on the obstacle state data at N historical times, where N is an integer greater than M, and the M historical times include at least one historical time adjacent to the target time period among the N historical times.

[0056] The following examples illustrate the example processes of the operations of the trajectory correction method in this embodiment.

[0057] Exemplarily, based on the obstacle state data at M historical times, the predicted obstacle state for the target time period is determined. The M historical times include at least one historical time adjacent to the target time period, for example, at least one historical time closest to the start time of the target time period. The start time of the target time period may be the current time.

[0058] In one example, the obstacle state data includes obstacle motion parameters and / or vehicle control parameters. At least one of the following driving parameters for the target time period may be determined based on the obstacle motion parameters and / or vehicle control parameters at M historical times as the predicted obstacle state: the obstacle driving speed, the obstacle driving acceleration, and the obstacle driving steering angle.

[0059] Determining the predicted motion state for the target time period based on the obstacle motion parameters and / or vehicle control parameters at M historical times can effectively improve the utilization rate of the obstacle state data, which is beneficial to ensuring the real-time performance and accuracy of obstacle trajectory prediction and providing reliable data support for autonomous driving control.

[0060] When the obstacle motion parameters and / or vehicle control parameters at M historical times meet the preset conditions, the predicted motion state of the obstacle is determined as the specified motion state. Based on the specified motion state, the predicted driving parameters for the target time period are determined as the predicted obstacle state. The obstacle motion parameters may include, for example, the obstacle speed parameter, the obstacle acceleration parameter, the obstacle steering angle parameter, etc. The vehicle control parameters may indicate the control state of the obstacle, for example, may indicate information such as the braking state, the turn signal state, and the steering motor torque of the obstacle.

[0061] Exemplarily, when the speed of an obstacle at M historical moments is greater than a preset speed threshold, the acceleration of the obstacle is greater than zero, the included angle between the steering angle of the obstacle and the lane line is less than a preset angle threshold, and the distance between the obstacle and the intersection is less than a preset distance threshold, it is determined that the predicted motion state of the obstacle is a straight-line motion state. Based on the straight-line motion state, it is determined that the lateral driving speed of the obstacle remains unchanged at any target moment within the target time period, and the longitudinal driving speed is determined according to the driving speed of the obstacle at the current moment and the driving acceleration of the obstacle. It can be understood that longitudinally can be the driving direction of the obstacle, and laterally can be the direction perpendicular to the driving direction of the obstacle.

[0062] Again exemplarily, when the acceleration of the obstacle at at least some of the M historical moments is less than zero or the brake light of the obstacle is on, it is determined that the predicted motion state of the obstacle is a decelerated motion state. Based on the decelerated motion state, the driving speed of the obstacle at any target moment within the target time period is determined as the predicted state of the obstacle.

[0063] Based on the straight-line motion state or the decelerated motion state, the reference obstacle position at at least one target moment within the target time period is determined. According to the reference obstacle position and the predicted obstacle position at at least one target moment indicated by the predicted obstacle trajectory, the trajectory correction parameter is determined.

[0064] When the predicted motion state of the obstacle is a straight-line motion state, the lateral position of the obstacle remains unchanged, and the longitudinal position of the obstacle is determined by the driving speed of the obstacle and the driving acceleration of the obstacle. For example, assuming that the position of the obstacle at the current moment is y(t0), the driving speed of the obstacle is v(t0), and the driving acceleration of the obstacle is a, the longitudinal position y(t0 + T) of the obstacle at any target moment (t0 + T) can be calculated using Equation (1) as the reference obstacle position:

[0065] y(t0 + T) = y(t0) + v(t0) * T + 1 / 2 * a * T 2 (1)

[0066] When the predicted motion state of the obstacle is a decelerated motion state, for example, assuming that the position of the obstacle at the current moment is y(t0), the driving speed of the obstacle is v(t0), and the driving acceleration of the obstacle is a. The driving speed v(t0 + T1) of the obstacle at the first target moment (t0 + T1) can be calculated using Equation (2), and the reference obstacle position y(t0 + T1) at the first target moment (t0 + T1) can be calculated using Equation (3):

[0067] v(t0 + T1) = v(t0) + a * T1 (2)

[0068] y(t0 + T1) = y(t0) + v(t0) * T1 + 1 / 2 * a * T1 2 (3)

[0069] The driving speed v(t0 + T1 + T2) of the obstacle at the second target time (t0 + T1 + T2) can be calculated using Equation (4), and the reference obstacle position y(t0 + T1 + T2) at the second target time (t0 + T1 + T2) can be calculated using Equation (5).

[0070] v(t0 + T1 + T2) = v(t0 + T1) (4)

[0071] y(t0 + T1 + T2) = y(t0 + T1) + v(t0 + T1)T2 (5)

[0072] For any target time, when the degree of difference between the reference obstacle position and the predicted obstacle position indicated by the predicted obstacle trajectory is greater than a preset threshold, it is determined that the state of the obstacle at at least some of the historical times adjacent to the current time may have changed. Due to possible deviations in the predicted obstacle position obtained from the obstacle state data based on N historical times, the reference obstacle position can be used to correct the predicted obstacle position.

[0073] Based on the reference obstacle position and the predicted obstacle position indicated by the predicted obstacle trajectory for at least one target time, a trajectory correction parameter is determined. For any target time, when the degree of difference between the reference obstacle position and the predicted obstacle position is greater than the preset threshold, the reference obstacle position can be used as the trajectory correction parameter to replace the predicted obstacle position with the reference obstacle position. Additionally, based on the reference obstacle position and the predicted obstacle position for at least one target time, a weighted obstacle position associated with the corresponding target time can be determined, and the weighted obstacle position can be used as the trajectory correction parameter to replace the predicted obstacle position with the weighted obstacle position.

[0074] Exemplarily, a reference distance between an obstacle and a target vehicle is determined according to a reference obstacle position based on a preset target time. A predicted distance between the obstacle and the target vehicle is determined according to a predicted obstacle position based on the same target time. In a case where a current distance between the obstacle and the target vehicle is less than a preset distance threshold (for example, the current distance between the obstacle and the target vehicle is less than 30 meters), a predicted distance between the obstacle and the target vehicle based on the preset target time is greater than the reference distance, and a difference between the predicted distance and the reference distance is greater than a preset difference threshold, it is determined that the state of the obstacle may have changed at at least some historical times adjacent to the current time, and it is determined that the obstacle has a deceleration intention at at least some historical times adjacent to the current time. Therefore, the reference obstacle position can be used as a trajectory correction parameter, and the predicted obstacle position can be replaced with the reference obstacle position to correct the predicted trajectory of the obstacle and obtain a corrected target obstacle trajectory.

[0075] In another example, the obstacle state data includes obstacle position parameters. When determining a predicted obstacle state according to the obstacle state data based on M historical times, an obstacle driving scenario can be determined according to the longitude and latitude coordinates indicated by the obstacle position parameters. According to the scenario type of the obstacle driving scenario and / or the obstacle position parameters, a driving parameter threshold associated with the obstacle driving scenario is determined. According to the driving parameter threshold, a predicted obstacle state for a target time period is determined.

[0076] Determining a predicted obstacle state for a target time period according to the scenario type of the obstacle driving scenario and / or the obstacle position parameters can effectively improve the accuracy of obstacle trajectory prediction and is beneficial to ensuring the safety and reliability of autonomous driving control.

[0077] Exemplarily, a high-precision map element corresponding to the longitude and latitude coordinates can be determined according to the longitude and latitude coordinates indicated by the obstacle position parameters. According to the high-precision map element, an obstacle driving scenario associated with the target time period is determined. The scenario type of the obstacle driving scenario can, for example, include a traffic intersection scenario and a speed limit section scenario. A driving parameter threshold associated with the obstacle driving scenario can be determined according to the scenario type of the obstacle driving scenario and / or the obstacle position parameters. The driving parameter threshold can, for example, include a driving speed threshold and a driving acceleration threshold. According to the driving parameter threshold, a predicted obstacle state for at least one target time in the target time period is determined.

[0078] When the obstacle driving scenario is a traffic intersection scenario, the traffic light status information in the traffic intersection scenario can be obtained. According to the traffic light status information, the target duration required for the traffic signal to change from green to red is determined. According to the target duration, the stop line of the traffic intersection, the current driving speed of the obstacle at the current moment, and the current position of the obstacle indicated by the obstacle position parameter, the driving acceleration threshold of the obstacle is determined to predict the obstacle state. The driving acceleration threshold can effectively ensure that the obstacle stops at the stop line of the traffic intersection from the current position of the obstacle within the target duration.

[0079] Exemplarily, based on the driving acceleration threshold, the reference obstacle position based on at least one target moment within the target time period can be determined. According to the reference obstacle position and the predicted obstacle position based on at least one target moment indicated by the predicted obstacle trajectory, the trajectory correction parameter is determined. For example, for any target moment, when the difference degree between the predicted obstacle position and the reference obstacle position is greater than the preset threshold, and the driving acceleration indicated by the predicted obstacle position is less than the driving acceleration threshold, the reference obstacle position is used as the trajectory correction parameter to replace the predicted obstacle position with the reference obstacle position.

[0080] When the obstacle driving scenario is a speed-limited section, the speed limited by the speed-limited section can be used as the driving speed threshold to obtain the predicted obstacle state. According to the driving speed threshold and the predicted obstacle trajectory for the target time period, the trajectory correction parameter is determined.

[0081] Exemplarily, based on the driving speed threshold, the reference obstacle position based on at least one target moment within the target time period can be determined. According to the reference obstacle position and the predicted obstacle position based on at least one target moment indicated by the predicted obstacle trajectory, the trajectory correction parameter is determined. For example, for any target moment, when the difference degree between the predicted obstacle position and the reference obstacle position is greater than the preset threshold, and the driving speed indicated by the predicted obstacle position is greater than the driving speed threshold, the reference obstacle position is used as the trajectory correction parameter to replace the predicted obstacle position with the reference obstacle position.

[0082] In another example, the obstacle state data includes obstacle attribute parameters, obstacle motion parameters, and driving lane parameters. When determining the predicted obstacle state based on the obstacle state data for M historical moments, the obstacle steering evaluation value for the target time period can be determined according to the obstacle attribute parameters, obstacle motion parameters, and driving lane parameters for M historical moments, and the predicted obstacle state can be determined according to the obstacle motion parameters and the obstacle steering evaluation value.

[0083] By determining the obstacle steering evaluation value for the target time period and predicting the obstacle state, it is beneficial to achieve reliable and effective obstacle trajectory prediction, and can provide credible decision support for driving assistance control.

[0084] Obstacle attribute parameters can indicate information such as obstacle type, obstacle width, and obstacle length. Driving lane parameters can indicate information such as lane width, lane line slope, and lateral distance between the obstacle and the lane line. Exemplarily, based on the obstacle attribute parameters, obstacle motion parameters, and driving lane parameters at M historical moments, a distance reference evaluation value and a steering angle reference evaluation value can be determined. Based on the distance reference evaluation value and the steering angle reference evaluation value, the obstacle steering evaluation value for the target time period is determined.

[0085] For example, the distance reference evaluation value g1(t0) at the current moment can be calculated using Equation (6)

[0086] g1(t0) = sigmoid[D / 2 - distance] (6)

[0087] The Sigmoid function can map data to the interval [-1, 1]. D represents the lane width, and distance represents the lateral distance between the vehicle head and the lane line. The distance can be calculated using Equation (7)

[0088] distance = ||d| - |L / 2 * sin(θ)|| (7)

[0089] d represents the lateral distance between the center point of the obstacle and the lane line, L represents the obstacle length, θ represents the difference between the obstacle steering angle and the lane line angle, and the lane line angle can be determined according to the lane line slope.

[0090] The steering angle reference evaluation value g2(t0) at the current moment can be calculated using Equation (8)

[0091] g2(t0) = flags * (|θ| + |speed * sin(θ)|) (8)

[0092] θ represents the difference between the obstacle steering angle and the lane line angle, speed represents the driving speed of the obstacle at the current moment, and flags is a flag value used to mark whether the obstacle turning angle direction is towards the inside of the lane. When the obstacle turning angle direction is towards the inside of the lane, the value of flags can be 1. When the obstacle turning angle direction is towards the outside of the lane, the value of flags can be -1.

[0093] The obstacle steering evaluation value for the target time period can be calculated using Equation (9)

[0094] G = x1 * g1(t0) + x2 * g2(t0) (9)

[0095] G represents the obstacle turning evaluation value for the target time period, and x1 and x2 respectively represent the preset weights corresponding to the distance reference evaluation value g1(t0) and the steering angle reference evaluation value g2(t0).

[0096] In the case where the obstacle turning evaluation value is less than the preset evaluation threshold, it is determined that the obstacle may continue to move forward along the current lane. In the case where the obstacle turning evaluation value is greater than or equal to the evaluation threshold, it is determined that the obstacle may switch lanes along the obstacle corner direction. According to the obstacle motion parameters and the obstacle turning evaluation value for the target time period, the predicted obstacle state is determined.

[0097] In another example, the obstacle state data may include the historical state data of multiple obstacles. When determining the predicted obstacle state according to the obstacle state data based on M historical moments, the historical motion features associated with each of the multiple obstacles can be determined according to the historical state data of the multiple obstacles based on M historical moments. According to the historical motion features associated with each obstacle, the spatio-temporal interaction features between the multiple obstacles are determined. For the target obstacle among the multiple obstacles, the predicted obstacle state is determined according to the spatio-temporal interaction features and the historical motion features associated with the target obstacle.

[0098] According to the spatio-temporal interaction features between the multiple obstacles, the predicted obstacle state for the target time period is determined. By considering the interaction relationship and interaction degree between different obstacles, the accuracy of obstacle trajectory prediction can be effectively guaranteed.

[0099] Exemplarily, the feature extraction sub-network in the trained trajectory prediction model can be used to determine the historical motion features associated with each of the multiple obstacles according to the historical state data of the multiple obstacles based on M historical moments. The attention sub-network in the trajectory prediction model is used to determine the spatial interaction features between the multiple obstacles according to the historical motion features associated with each obstacle. The recurrent neural sub-network in the trajectory prediction model is used to obtain the spatio-temporal interaction features between the multiple obstacles according to the spatial interaction features between the multiple obstacles. The trajectory prediction model can be, for example, an Interaction Prediction Network model based on the attention mechanism.

[0100] According to the spatio-temporal interaction features and the historical motion features associated with the target obstacle, the predicted motion state for the target time period associated with the target obstacle is determined. For example, the spatio-temporal interaction features and the historical motion features associated with the target obstacle are encoded and decoded to obtain the predicted motion state of the target obstacle.

[0101] Based on the predicted motion state, determine the reference obstacle position based on at least one target moment within the target time period. Determine the trajectory correction parameter according to the reference obstacle position and the predicted obstacle position based on at least one target moment indicated by the predicted obstacle trajectory. Based on the trajectory correction parameter, correct the predicted obstacle trajectory to obtain the corrected target obstacle trajectory.

[0102] Exemplarily, a trained trajectory prediction model can be used to determine the predicted obstacle trajectory for the target time period according to the obstacle state data based on N historical moments. The obstacle state data based on N historical moments can be high-dimensional feature data for obstacle trajectory prediction. The high-dimensional feature data can include, for example, obstacle attribute parameters, obstacle motion parameters, driving lane parameters, multi-obstacle interaction feature parameters, obstacle space interaction feature parameters, etc.

[0103] When performing obstacle trajectory prediction for the target time period according to the high-dimensional feature data based on N historical moments, the target time period can include at least one target moment starting from the current moment, and the N historical moments can include at least one historical moment ending at the current moment. When performing obstacle trajectory prediction according to the high-dimensional feature data based on N historical moments, the trajectory prediction model may be difficult to capture the obstacle state change information within a short time, and the predicted obstacle trajectory output by the trajectory prediction model may deviate from the actual obstacle trajectory.

[0104] For example, when the obstacle state data changes at m historical moments adjacent to the current moment, the trajectory prediction model may be difficult to capture the obstacle state change information at the m historical moments, where m is a positive integer less than N. Therefore, the obstacle trajectory can be corrected according to the obstacle state data of the partial historical moments closest to the current moment among the N historical moments.

[0105] Determine the predicted obstacle state for the target time period according to the obstacle state data based on M historical moments, and correct the predicted obstacle trajectory according to the predicted obstacle state to obtain the corrected target obstacle trajectory, where the M historical moments are the partial historical moments closest to the current moment among the N historical moments. Determining the predicted obstacle state for the target time period according to the obstacle state data based on M historical moments can effectively improve the utilization rate of the obstacle state data, can effectively capture and utilize the obstacle motion change information within a short time, is beneficial to improving the accuracy of obstacle trajectory prediction, can achieve more refined autonomous driving control, and is beneficial to ensuring the safe driving of autonomous vehicles.

[0106] Figure 4 A schematic diagram schematically shows the trajectory correction process according to an embodiment of the present disclosure.

[0107] As Figure 4 shown, in the trajectory correction process 400, based on the obstacle state data 430 at M historical moments, the predicted obstacle state 440 for the target time period is determined. According to the predicted obstacle state 440 and the predicted obstacle trajectory 420 for the target time period, the trajectory correction parameter 450 is determined.

[0108] The predicted obstacle trajectory 420 is obtained based on the obstacle state data 410 at N historical moments, where N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments. Based on the trajectory correction parameter 450, the predicted obstacle trajectory 420 is corrected to obtain the corrected target obstacle trajectory 460.

[0109] According to the obstacle state data of some historical moments adjacent to the target time period, the predicted obstacle state for the target time period is determined. Using the predicted obstacle state, the pre-determined predicted obstacle trajectory is corrected to obtain the corrected target obstacle trajectory. It is beneficial to achieve reliable and effective obstacle trajectory prediction, can effectively improve the accuracy of obstacle trajectory prediction, can provide credible decision support for driving assistance control, and is beneficial to achieve safe and reliable autonomous driving control.

[0110] Figure 5 Schematically shows a block diagram of a trajectory correction device according to an embodiment of the present disclosure.

[0111] As Figure 5 shown, the trajectory correction device 500 according to an embodiment of the present disclosure includes, for example, a first processing module 510, a second processing module 520, and a third processing module 530.

[0112] The first processing module 510 is configured to determine the predicted obstacle state for the target time period in response to the obstacle state data at M historical moments, where M is an integer greater than 0; the second processing module 520 is configured to determine the trajectory correction parameter according to the predicted obstacle state and the predicted obstacle trajectory for the target time period; and the third processing module 530 is configured to correct the predicted obstacle trajectory based on the trajectory correction parameter to obtain the corrected target obstacle trajectory. The predicted obstacle trajectory is obtained based on the obstacle state data at N historical moments, where N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0113] In an embodiment of the present disclosure, in response to obstacle state data based on M historical moments, a predicted obstacle state for a target time period is determined, where M is an integer greater than 0. According to the predicted obstacle state and the predicted obstacle trajectory for the target time period, a trajectory correction parameter is determined, and based on the trajectory correction parameter, the predicted obstacle trajectory is corrected to obtain a corrected target obstacle trajectory. The predicted obstacle trajectory is obtained based on obstacle state data of N historical moments, where N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments.

[0114] Based on obstacle state data of at least one historical moment adjacent to the target time period, a predicted obstacle state associated with the target time period is determined, and based on the predicted obstacle state and a pre-determined predicted obstacle trajectory, a trajectory correction parameter is determined. By improving the utilization rate of obstacle state data, the accuracy of obstacle trajectory prediction can be effectively improved, the precision of autonomous driving control can be effectively enhanced, credible decision support can be provided for driving assistance control, and it is beneficial to ensure the safe driving of autonomous vehicles.

[0115] According to an embodiment of the present disclosure, the obstacle state data includes obstacle motion parameters and / or vehicle control parameters; the first processing module includes: a first processing sub-module, configured to determine at least one of the following driving parameters for the target time period based on the obstacle motion parameters and / or vehicle control parameters of M historical moments as the predicted obstacle state: the obstacle driving speed, the obstacle driving acceleration, and the obstacle driving steering angle.

[0116] According to an embodiment of the present disclosure, the obstacle state data includes obstacle position parameters; the first processing module includes: a second processing sub-module, configured to determine the obstacle driving scenario according to the longitude and latitude coordinates indicated by the obstacle position parameters; a third processing sub-module, configured to determine a driving parameter threshold associated with the obstacle driving scenario according to the scenario type of the obstacle driving scenario and / or the obstacle position parameters; and a fourth processing sub-module, configured to determine the predicted obstacle state for the target time period according to the driving parameter threshold.

[0117] According to an embodiment of the present disclosure, the obstacle state data includes obstacle attribute parameters, obstacle motion parameters, and driving lane parameters; the first processing module includes: a fifth processing sub-module, configured to determine an obstacle steering evaluation value for the target time period based on the obstacle attribute parameters, obstacle motion parameters, and driving lane parameters of M historical moments; and a sixth processing sub-module, configured to determine the predicted obstacle state according to the obstacle motion parameters and the obstacle steering evaluation value.

[0118] According to an embodiment of the present disclosure, the obstacle state data includes historical state data of multiple obstacles; the first processing module includes: a seventh processing sub-module, configured to determine historical motion features associated with each of the multiple obstacles according to the historical state data of the multiple obstacles based on M historical moments; an eighth processing sub-module, configured to determine spatio-temporal interaction features between the multiple obstacles according to the historical motion features associated with each obstacle; and a ninth processing sub-module, configured to determine a predicted obstacle state for a target obstacle among the multiple obstacles according to the spatio-temporal interaction features and the historical motion features associated with the target obstacle.

[0119] According to an embodiment of the present disclosure, the second processing module includes: a tenth processing sub-module, configured to determine a reference obstacle position based on at least one target moment within a target time period according to the predicted obstacle state; and an eleventh processing sub-module, configured to determine a trajectory correction parameter according to the reference obstacle position and the predicted obstacle position indicated by the predicted obstacle trajectory based on at least one target moment.

[0120] According to an embodiment of the present disclosure, the eleventh processing sub-module includes: a first processing unit, configured to, for any target moment, when the degree of difference between the reference obstacle position and the predicted obstacle position is greater than a preset threshold, use the reference obstacle position as the trajectory correction parameter; or a second processing unit, configured to, for any target moment, determine a weighted obstacle position based on the corresponding target moment according to the reference obstacle position and the predicted obstacle position, and use it as the trajectory correction parameter.

[0121] According to an embodiment of the present disclosure, the third processing module includes a twelfth processing sub-module, configured to use the trajectory correction parameter to replace the predicted obstacle position based on the corresponding target moment in the predicted obstacle trajectory, and obtain a corrected target obstacle trajectory.

[0122] According to an embodiment of the present disclosure, the apparatus further includes a fourth processing module, configured to generate a vehicle control instruction based on the target obstacle trajectory; and send the vehicle control instruction to a vehicle control terminal to control the vehicle to travel based on the vehicle control instruction.

[0123] It should be noted that in the technical solution of the present disclosure, the processing of information collection, storage, use, processing, transmission, provision, and disclosure all comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0124] According to an embodiment of the present disclosure, there is provided an electronic device, including: at least one processor and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the above-mentioned trajectory correction method.

[0125] According to an embodiment of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the above-mentioned trajectory correction method.

[0126] According to an embodiment of the present disclosure, there is provided a computer program product including a computer program / instructions, which when executed by a processor, implement the above-mentioned trajectory correction method.

[0127] According to an embodiment of the present disclosure, there is further provided a cloud control platform, which, for example, includes the above-mentioned electronic device.

[0128] According to an embodiment of the present disclosure, there is further provided an autonomous vehicle, which, for example, includes an electronic device. The electronic device includes at least one processor and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, and the instructions, when executed by the at least one processor, enable the at least one processor to execute the above-mentioned model training method or execute the above-mentioned trajectory prediction method. Exemplarily, the electronic device according to an embodiment of the present disclosure is, for example, similar to Figure 6 the electronic device shown.

[0129] Figure 6 Schematically shows a block diagram of an electronic device for executing a trajectory correction method according to an embodiment of the present disclosure.

[0130] Figure 6 Shows a schematic block diagram of an example electronic device 600 that can be used to implement an embodiment of the present disclosure. The electronic device 600 is intended to represent various forms of digital computers, such as, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0131] As Figure 6 shown, the device 600 includes a computing unit 601, which can execute various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other through a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0132] Multiple components in device 600 are connected to I / O interface 605, including: input unit 606, such as a keyboard, mouse, etc.; output unit 607, such as various types of displays, speakers, etc.; storage unit 608, such as a disk, optical disc, etc.; and communication unit 609, such as a network card, modem, wireless communication transceiver, etc. Communication unit 609 allows device 600 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0133] Computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of computing unit 601 include but are not limited to a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running deep learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Computing unit 601 executes the various methods and processes described above, such as the trajectory correction method. For example, in some embodiments, the trajectory correction method can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed onto device 600 via ROM 602 and / or communication unit 609. When the computer program is loaded into RAM 603 and executed by computing unit 601, one or more steps of the trajectory correction method described above can be executed. Alternatively, in other embodiments, computing unit 601 can be configured to execute the trajectory correction method by any other suitable means (e.g., by means of firmware).

[0134] The various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGA), application specific integrated circuits (ASIC), application specific standard products (ASSP), systems on a chip (SOC), complex programmable logic devices (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special or general-purpose programmable processor, and can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0135] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable trajectory correction devices, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowchart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, executed partially on the machine and partially on a remote machine as an independent software package, or executed entirely on a remote machine or server.

[0136] In the context of the present disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0137] In order to provide interaction with an object, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the object; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the object can provide input to the computer. Other kinds of devices can also be used to provide interaction with the object; for example, the feedback provided to the object can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the object can be received in any form (including acoustic input, voice input, or tactile input).

[0138] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., an object computer having a graphical object interface or a web browser through which an object can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.

[0139] A computer system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The client - server relationship is created by computer programs running on respective computers and having a client - server relationship with each other. The server can be a cloud server, a server of a distributed system, or a server incorporating blockchain.

[0140] It should be understood that the various forms of the processes shown above can be reordered, steps added, or steps deleted. For example, the steps recited in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitation is imposed herein.

[0141] The above - described specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub - combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. A trajectory correction method, comprising: Determining a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments, where M is an integer greater than 0; Determining a trajectory correction parameter according to the predicted obstacle state and a predicted obstacle trajectory for the target time period; And Based on the trajectory correction parameter, correcting the predicted obstacle trajectory to obtain a corrected target obstacle trajectory, Wherein the predicted obstacle trajectory is obtained based on obstacle state data of N historical moments, N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments; Wherein, the determining a trajectory correction parameter according to the predicted obstacle state and a predicted obstacle trajectory for the target time period includes: Determining a reference obstacle position based on at least one target moment within the target time period according to the predicted obstacle state; and Determining the trajectory correction parameter according to the reference obstacle position and a predicted obstacle position indicated by the predicted obstacle trajectory based on the at least one target moment, including: For any target moment, when the degree of difference between the reference obstacle position and the predicted obstacle position is greater than a preset threshold, using the reference obstacle position as the trajectory correction parameter; or For any target moment, determining a weighted obstacle position based on the corresponding target moment according to the reference obstacle position and the predicted obstacle position to be used as the trajectory correction parameter.

2. The method according to claim 1, wherein The obstacle state data includes obstacle motion parameters and / or vehicle control parameters; The determining a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments includes: Determining at least one of the following driving parameters for the target time period according to the obstacle motion parameters and / or the vehicle control parameters based on the M historical moments to be used as the predicted obstacle state: Obstacle driving speed, obstacle driving acceleration, and obstacle driving steering angle.

3. The method according to claim 1, wherein The obstacle state data includes obstacle position parameters; The determining a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments includes: Determining an obstacle driving scenario according to the longitude and latitude coordinates indicated by the obstacle position parameters; Determining a driving parameter threshold associated with the obstacle driving scenario according to the scenario type of the obstacle driving scenario and / or the obstacle position parameters; and Determining a predicted obstacle state for the target time period according to the driving parameter threshold.

4. The method according to claim 1, wherein The obstacle state data includes obstacle attribute parameters, obstacle motion parameters, and driving lane parameters; The determining a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments includes: Determine an obstacle steering evaluation value for the target time period based on the obstacle attribute parameters, the obstacle motion parameters, and the driving lane parameters for the M historical moments; And Determine the predicted obstacle state based on the obstacle motion parameters and the obstacle steering evaluation value.

5. The method according to claim 1, wherein The obstacle state data includes historical state data of multiple obstacles; The determining the predicted obstacle state for the target time period in response to the obstacle state data based on M historical moments includes: Determine a historical motion feature associated with each obstacle among the multiple obstacles according to the historical state data of the multiple obstacles based on the M historical moments; Determine the spatio-temporal interaction feature between the multiple obstacles according to the historical motion feature associated with each obstacle; and For a target obstacle among the multiple obstacles, determine the predicted obstacle state according to the spatio-temporal interaction feature and the historical motion feature associated with the target obstacle.

6. The method according to claim 1, wherein, The modifying the predicted obstacle trajectory based on the trajectory correction parameter to obtain a modified target obstacle trajectory includes Using the trajectory correction parameter to replace the predicted obstacle position based on the corresponding target moment in the predicted obstacle trajectory to obtain the modified target obstacle trajectory.

7. The method according to claim 1, further comprising: Generating a vehicle control instruction based on the target obstacle trajectory; And Sending the vehicle control instruction to a vehicle control terminal so as to control the vehicle to travel based on the vehicle control instruction.

8. A trajectory correction device, comprising: A first processing module, configured to determine a predicted obstacle state for a target time period in response to obstacle state data based on M historical moments, where M is an integer greater than 0; A second processing module, configured to determine a trajectory correction parameter according to the predicted obstacle state and the predicted obstacle trajectory for the target time period; And A third processing module, configured to modify the predicted obstacle trajectory based on the trajectory correction parameter to obtain a modified target obstacle trajectory, wherein the predicted obstacle trajectory is obtained based on the obstacle state data of N historical moments, N is an integer greater than M, and the M historical moments include at least one historical moment adjacent to the target time period among the N historical moments; wherein the second processing module includes: A tenth processing sub-module, configured to determine a reference obstacle position based on at least one target moment within the target time period according to the predicted obstacle state; and An eleventh processing sub-module, configured to determine the trajectory correction parameter according to the reference obstacle position and the predicted obstacle position based on the at least one target moment indicated by the predicted obstacle trajectory; wherein the eleventh processing sub-module includes: A first processing unit, configured to, for any target moment, in the case that the difference degree between the reference obstacle position and the predicted obstacle position is greater than a preset threshold, use the reference obstacle position as the trajectory correction parameter; or A second processing unit, configured to determine a weighted obstacle position based on a corresponding target time according to the reference obstacle position and the predicted obstacle position, as the trajectory correction parameter for any target time.

9. The apparatus according to claim 8, wherein, The obstacle state data includes obstacle motion parameters and / or vehicle control parameters; The first processing module includes: A first processing sub-module, configured to determine at least one of the following driving parameters for the target time period according to the obstacle motion parameters and / or the vehicle control parameters based on the M historical times, as the predicted obstacle state: Obstacle driving speed, obstacle driving acceleration, and obstacle driving steering angle.

10. The device according to claim 8, wherein, The obstacle state data includes obstacle position parameters; The first processing module includes: A second processing sub-module, configured to determine an obstacle driving scenario according to the longitude and latitude coordinates indicated by the obstacle position parameters; A third processing sub-module, configured to determine a driving parameter threshold associated with the obstacle driving scenario according to the scenario type of the obstacle driving scenario and / or the obstacle position parameters; and A fourth processing sub-module, configured to determine the predicted obstacle state for the target time period according to the driving parameter threshold.

11. The device according to claim 8, wherein, The obstacle state data includes obstacle attribute parameters, obstacle motion parameters, and driving lane parameters; The first processing module includes: A fifth processing sub-module, configured to determine an obstacle steering evaluation value for the target time period according to the obstacle attribute parameters, the obstacle motion parameters, and the driving lane parameters based on the M historical times; And A sixth processing sub-module, configured to determine the predicted obstacle state according to the obstacle motion parameters and the obstacle steering evaluation value.

12. The apparatus according to claim 8, wherein, The obstacle state data includes historical state data of multiple obstacles; the first processing module includes: A seventh processing sub-module, configured to determine a historical motion feature associated with each obstacle among the multiple obstacles according to the historical state data of the multiple obstacles based on the M historical times; An eighth processing sub-module, configured to determine a spatio-temporal interaction feature between the multiple obstacles according to the historical motion features associated with each obstacle; and A ninth processing sub-module, configured to determine the predicted obstacle state for a target obstacle among the multiple obstacles according to the spatio-temporal interaction feature and the historical motion feature associated with the target obstacle.

13. The apparatus according to claim 8, wherein, The third processing module includes A twelfth processing sub-module, configured to replace the predicted obstacle position based on the corresponding target time in the predicted obstacle trajectory with the trajectory correction parameter to obtain the corrected target obstacle trajectory.

14. The apparatus according to claim 8, further comprising a fourth processing module, configured to: Generate a vehicle control instruction based on the target obstacle trajectory; and Send the vehicle control instruction to a vehicle control terminal to control the vehicle to travel based on the vehicle control instruction.

15. An electronic device, comprising: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method according to any one of claims 1 to 7.

16. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are for causing the computer to execute the method according to any one of claims 1 to 7.

17. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

18. A cloud control platform, comprising the electronic device according to claim 15.

19. An autonomous vehicle, comprising the electronic device according to claim 15.

Images