Trajectory Planning Method and Device, Server, Computer-Readable Storage Medium

Abstract

The present invention discloses a trajectory planning method and apparatus, a server, and a computer-readable storage medium. The trajectory planning method includes: obtaining first surrounding environment information of a first target vehicle; determining at least one second target having an interaction relationship with the first target vehicle among other targets according to the first surrounding environment information to form at least one target pair; for each target pair, predicting at least one first trajectory of the first target vehicle and at least one second trajectory of the second target; predicting a third trajectory and its occurrence probability of the first target vehicle according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair; and selecting an optimal trajectory of the first target vehicle according to the occurrence probabilities of the respective third trajectories. Using the above technical solution can reduce the computational amount required for trajectory prediction to improve the response speed of a driverless vehicle.

Description

Technical Field

[0001] The present invention relates to the technical field of driverless technology, and more specifically, to a trajectory planning method and apparatus, a server, and a computer-readable storage medium. Background Art

[0002] With the gradual maturity of driverless technology, driverless technology is applied to various scenarios. Driverless technology can replace manual control of vehicle driving and make judgments on the behaviors of surrounding targets according to external situations to plan the behaviors of the vehicle.

[0003] In the prior art, when performing trajectory planning for a driverless vehicle, it is usually to collect information of the driverless vehicle and information of surrounding targets, predict the trajectories of surrounding targets based on the obtained information, and plan the trajectory of the driverless vehicle accordingly.

[0004] However, in the prior art, when performing trajectory prediction, it is necessary to predict the trajectories of all surrounding targets at the same time, consuming a large amount of computing power, which will cause a certain delay for the driverless vehicle from obtaining information to performing operations, resulting in unknown safety hazards for the driverless vehicle on the road. Summary of the Invention

[0005] The technical problem solved by the present invention is how to reduce the computing power required for trajectory prediction when planning the trajectory of a driverless vehicle to improve the response speed of the driverless vehicle.

[0006] To solve the above technical problem, an embodiment of the present invention provides a trajectory planning method, where the trajectory planning method includes: obtaining first surrounding environment information of a first target vehicle, where the first surrounding environment information includes action information of other targets; determining at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information to form at least one target pair, and each target pair includes the first target vehicle and a second target; for each target pair, predicting at least one first trajectory of the first target vehicle according to the first surrounding environment information, and predicting at least one second trajectory of the second target according to second surrounding environment information of the second target; predicting a third trajectory of the first target vehicle and its occurrence probability according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair, where the occurrence probability represents the probability that the first target vehicle travels according to the third trajectory; and selecting an optimal trajectory of the first target vehicle according to the occurrence probabilities of the respective third trajectories.

[0007] Optionally, determining at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information includes: inputting the first surrounding environment information into a collision model to obtain at least one second target having an interaction relationship with the first target vehicle.

[0008] Optionally, predicting a third trajectory and its occurrence probability of the first target vehicle according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair includes: inputting the first trajectory and the second trajectory into a trajectory prediction model to obtain the third trajectory and its occurrence probability of the first target vehicle.

[0009] Optionally, predicting at least one first trajectory of the first target vehicle according to the first surrounding environment information, and predicting at least one second trajectory of the second target according to the second surrounding environment information of the second target includes: predicting the first trajectory using the first surrounding environment information after removing redundant data; predicting the second trajectory using the second surrounding environment information after removing redundant data.

[0010] Optionally, obtaining the first surrounding environment information of the first target vehicle includes: combining the obtained map information, action information of surrounding targets, and signal light information to form the first surrounding environment information.

[0011] Optionally, the method for obtaining the map information includes: obtaining map data pre-stored in a database, and determining the map information of the area where the vehicle is located according to the positioning information of the first target vehicle.

[0012] Optionally, before selecting the optimal trajectory of the first target vehicle according to the occurrence probability of each third trajectory, it includes: adjusting the occurrence probability of each third trajectory according to safety constraint conditions to reduce the occurrence probability of the third trajectory that does not meet the safety constraint conditions.

[0013] Optionally, after selecting the optimal trajectory of the first target vehicle according to the occurrence probability of each third trajectory, it includes: adjusting the optimal trajectory according to driving constraint conditions, where the driving constraint conditions are used to limit the speed and turning angle of the first target vehicle.

[0014] Optionally, before determining at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information, it includes: performing target recognition on the other targets to determine the type label of the other targets; adding the type label of the other targets to the first surrounding environment information.

[0015] An embodiment of the present invention also discloses a trajectory planning device, which includes: an acquisition module for acquiring first surrounding environment information of a first target vehicle, where the first surrounding environment information includes action information of other targets; a target pair generation module for determining at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information to form at least one target pair, and each target pair includes the first target vehicle and a second target; a first prediction module for, for each target pair, predicting at least one first trajectory of the first target vehicle according to the first surrounding environment information and predicting at least one second trajectory of the second target according to the second surrounding environment information of the second target; a second prediction module for predicting a third trajectory and its occurrence probability of the first target vehicle according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair, where the occurrence probability represents the probability that the first target vehicle travels according to the third trajectory; and a selection module for selecting an optimal trajectory of the first target vehicle according to the occurrence probabilities of the respective third trajectories.

[0016] An embodiment of the present invention also discloses a server, which includes a memory and a processor, and a computer program that can run on the processor is stored on the memory, and when the computer program is run by the processor, it executes the steps of any one of the above-mentioned trajectory planning methods.

[0017] An embodiment of the present invention also discloses a computer-readable storage medium, on which a computer program is stored. The computer-readable storage medium is a non-volatile storage medium or a non-transient storage medium, and when the computer program is run by the processor, it executes the steps of any one of the above-mentioned trajectory planning methods.

[0018] Compared with the prior art, the technical solution of the embodiment of the present invention has the following beneficial effects:

[0019] The present invention provides a trajectory planning method. By obtaining the first surrounding environment information of the first target vehicle, the action information of other targets is obtained, and at least one second target having an interaction relationship with the first target vehicle is determined from all other targets according to the first surrounding environment information to form at least one target pair. Screening out the second targets that have no interaction relationship with the first target vehicle can simplify the scenario where the first target vehicle is located into a more easily analyzable scenario, reducing the computing power consumption during trajectory prediction. Then, for each target pair, the first trajectory of the first target vehicle is predicted according to the first surrounding environment information, and the second trajectory of the second target is predicted according to the second surrounding environment information of the second target. Predicting the trajectories of the second target pairs can use all possible trajectories of the second target as a reference for the first target vehicle to perform the third trajectory planning, making the trajectory planning of the first target vehicle more accurate. Then, according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair, the third trajectory of the first target vehicle and its occurrence probability are predicted, and the optimal trajectory of the first target vehicle is selected according to the occurrence probability of each third trajectory. Using all the predicted trajectories of the first target vehicle and the second target as a reference for predicting the third trajectory enables the third trajectory to meet the driving requirements in various situations, making the trajectory planning more scientific and accurate.

[0020] Further, adjusting the occurrence probability of each third trajectory according to the safety constraint conditions can reduce the occurrence probability of the third trajectories that do not meet the safety constraint conditions, making the optimal trajectory selected by the first target vehicle safer.

[0021] Further, classifying each second target can predict the second trajectory according to different categories of second targets, making the second trajectory more in line with the movement laws of each second target. Description of the Drawings

[0022] Figure 1 is the overall flowchart of a trajectory planning method provided by an embodiment of the present invention;

[0023] Figure 2 is a schematic diagram of a trajectory planning scenario provided by an embodiment of the present invention;

[0024] Figure 3 is a schematic diagram of another trajectory planning scenario provided by an embodiment of the present invention;

[0025] Figure 4 is a schematic diagram of yet another trajectory planning scenario provided by an embodiment of the present invention;

[0026] Figure 5 is a schematic diagram of the structure of a trajectory planning device provided by an embodiment of the present invention. Detailed Embodiments

[0027] As described in the background art, with the gradual maturity of driverless technology, driverless technology is applied to various scenarios. The driverless technology can replace manual control of vehicle driving and make judgments on the behaviors of surrounding targets according to external situations to plan the behaviors of the vehicle. The prior art collects information of driverless vehicles and information of surrounding targets, predicts the trajectories of surrounding targets based on the obtained information, and plans the trajectories of driverless vehicles accordingly. However, when predicting trajectories in the prior art, it is necessary to predict the trajectories of all surrounding targets simultaneously, which consumes a large amount of computing power. This will cause a certain delay for the driverless vehicle from obtaining information to executing operations, resulting in unknown safety hazards for the driverless vehicle on the road.

[0028] In an embodiment of the present invention, by obtaining the first surrounding environment information of the first target vehicle to obtain the action information of other targets, and determining at least one second target having an interaction relationship with the first target vehicle among all other targets according to the first surrounding environment information to form at least one target pair. Screening out the second targets that have no interaction relationship with the first target vehicle can simplify the scenario where the first target vehicle is located into a scenario that is easier to analyze, reducing the computing power consumption when predicting trajectories. Then, for each target pair, predict the first trajectory of the first target vehicle according to the first surrounding environment information, and predict the second trajectory of the second target according to the second surrounding environment information of the second target. Predicting the trajectories of the second target pairs can use all possible trajectories of the second target as a reference for the first target vehicle to perform the third trajectory planning, making the trajectory planning of the first target vehicle more accurate. Then, predict the third trajectory of the first target vehicle and its occurrence probability according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair, and select the optimal trajectory of the first target vehicle according to the occurrence probability of each third trajectory. Using all the predicted trajectories of the first target vehicle and the second target as a reference for predicting the third trajectory enables the third trajectory to meet the driving requirements in various situations, making the trajectory planning more scientific and accurate.

[0029] Further, adjusting the occurrence probability of each third trajectory according to the safety constraint conditions can reduce the occurrence probability of the third trajectories that do not meet the safety constraint conditions, making the optimal trajectory selected by the first target vehicle safer.

[0030] Further, classifying each second target can predict the second trajectory according to different types of second targets, making the second trajectory more in line with the motion laws of each second target.

[0031] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of specific embodiments of the present invention with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0032] Figure 1 is the overall flowchart of a trajectory planning method provided by an embodiment of the present invention.

[0033] In specific implementation, the trajectory planning method described in the following steps 101 to 105 can be used in a server. The above steps can be specifically executed by the server, or by a chip with data processing capabilities in the server, or by a chip module including a chip with data processing capabilities in the server. In a specific embodiment, each step of the trajectory planning method can be executed by the server.

[0034] Specifically, as Figure 1 shown, the trajectory planning method may include the following steps:

[0035] In step 101, obtain the first surrounding environment information of the first target vehicle, where the first surrounding environment information includes the action information of other targets;

[0036] In step 102, determine at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information, and form at least one target pair, where each target pair includes the first target vehicle and a second target;

[0037] In step 103, for each target pair, predict at least one first trajectory of the first target vehicle according to the first surrounding environment information, and predict at least one second trajectory of the second target according to the second surrounding environment information of the second target;

[0038] In step 104, predict the third trajectory of the first target vehicle and its occurrence probability according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair, where the occurrence probability represents the probability that the first target vehicle travels according to the third trajectory;

[0039] In step 105, select the optimal trajectory of the first target vehicle according to the occurrence probabilities of each third trajectory.

[0040] So far, the optimal trajectory of the first target vehicle has been selected, and the first target vehicle can perform driverless driving according to the optimal trajectory.

[0041] In the specific implementation of step 101, the first surrounding environment information collected in real time by the sensor is obtained. The first surrounding environment information may include the action information of other targets, and the action information may include the position, speed, and direction of other targets. The first surrounding environment information may also include map information and traffic signal information.

[0042] In the specific implementation, the map data pre-stored in the database can be obtained, and the map information of the area where the vehicle is located can be determined according to the positioning information of the first target vehicle. The map information may include the lane line information near the current position of the vehicle, so as to facilitate controlling the first target vehicle to drive in the correct lane.

[0043] Furthermore, the traffic signal information can be obtained according to the photographing device on the first target vehicle to determine the current traffic signal situation, so as to plan the vehicle trajectory according to the indication of the traffic signal.

[0044] In a non-limiting embodiment, other targets are first identified to determine the type labels of other targets, and the type labels of other targets are added to the first surrounding environment information. Specifically, the type labels may include pedestrians, non-motor vehicles, and automobiles. Different types of other targets have different movement laws. By classifying other targets, accurate prediction can be made according to the movement laws of different target types when predicting the trajectories of targets subsequently.

[0045] In the specific implementation, the first surrounding environment information and the state information of the first target vehicle can be input into the collision model, and at least one second target having an interaction relationship with the first target vehicle can be obtained. At least one target pair is formed by the first target vehicle and the second target. The state information of the first target vehicle represents the position, speed, and direction of the first target vehicle. That the second target has an interaction relationship with the first target vehicle means that the second target may affect the driving of the first target vehicle. Specifically, the collision model may be a mathematical model.

[0046] It should be noted that the second target may be any movable object such as a vehicle or a pedestrian that can be implemented, and the present application does not limit this.

[0047] In a non-limiting embodiment, redundant data can be removed from the first surrounding environment information and the second surrounding environment information. Specifically, other objects that have no interaction relationship with the first target vehicle are removed from the first surrounding environment information, and other objects that have no interaction relationship with the second object in the object pair are removed from the second surrounding environment information. Then, with the two objects in the object pair as the main perspectives respectively, the range of the map and the orientation of the vehicle are re-planned. By reducing the range of the map and removing other objects that have no interaction relationship, redundant data in the scenario can be reduced, and the computing efficiency of the model can be improved.

[0048] In a specific implementation, the first surrounding environment information after removing redundant data and the second surrounding environment information after removing redundant data are input into the object pair trajectory prediction model, so as to predict the first trajectory of the first target vehicle within a preset time according to the first surrounding environment information, and predict the second trajectory of the second object within a preset time according to the second surrounding environment information. Specifically, the object pair trajectory prediction model can be a pre-trained neural network model, and the object pair trajectory prediction model can predict the trajectory of the object.

[0049] In a non-limiting embodiment, the first trajectory and the second trajectory output by the object pair trajectory prediction model are input into the trajectory prediction model to obtain the third trajectory of the first target vehicle and its occurrence probability.

[0050] In a specific implementation, the object pair trajectory prediction model can output the first trajectory and its occurrence probability and the second trajectory and its occurrence probability. The occurrence probability of the first trajectory refers to the probability that the first target vehicle travels along the first trajectory, and the occurrence probability of the second trajectory refers to the probability that the second object travels along the second trajectory. The trajectory prediction model plans the third trajectory according to the first trajectory and the second trajectory to obtain the third trajectory and its occurrence probability.

[0051] Furthermore, the trajectory prediction model can be a pre-trained neural network model, and the training set of the trajectory prediction model can include the real trajectories of vehicles, making the prediction of the trajectory prediction model more accurate.

[0052] In a non-limiting embodiment, the occurrence probabilities of each third trajectory are sorted, and the third trajectory with the highest occurrence probability is selected as the optimal trajectory of the first target vehicle.

[0053] In a specific implementation, before sorting the occurrence probabilities of each third trajectory, the occurrence probabilities of each third trajectory are adjusted according to safety constraint conditions to reduce the occurrence probabilities of the third trajectories that do not meet the safety constraint conditions. Specifically, the safety constraint conditions may be related to requiring the first target vehicle to maintain a safe distance from other targets. When the first target vehicle collides with the second target or surrounding obstacles, or when the distance between the first target vehicle and the second target or surrounding obstacles is too close, the occurrence probability of the third trajectory is reduced to ensure that the finally selected optimal trajectory is safe and reliable.

[0054] In a non-limiting embodiment, the target pair trajectory prediction model and the output of the trajectory prediction model can be a complete trajectory, or the position, speed, and direction of the target at specific time points within a preset time. For example, the output of the model can be a complete trajectory in the next 1 to 2 seconds, or the position, speed, and direction of the first target vehicle or the second target at the 1st second and the 2nd second. When the output of the model is the position, speed, and direction of the target at a specific time point, an interpolation method can be used according to the output of the model to generate a smooth curve to obtain a complete trajectory.

[0055] In a specific implementation, after obtaining the optimal trajectory, the acceleration, speed, and turning angle of the first target vehicle at each moment are calculated according to the optimal trajectory to obtain the specific operations that the first target vehicle needs to perform. The speed of the first target vehicle can be calculated based on the current speed and acceleration.

[0056] Furthermore, after obtaining the optimal trajectory, the optimal trajectory can be adjusted according to driving constraint conditions. Specifically, the speed and turning angle of the first target vehicle are restricted, and the change frequencies of the speed and turning angle of the first target vehicle are controlled within a first preset range to avoid frequent speed changes and turning of the first target vehicle, making the obtained optimal trajectory more in line with normal human driving habits; and controlling the speed and turning angle of the first target vehicle not to exceed a second preset range, making the obtained optimal trajectory safer.

[0057] In this embodiment, by analyzing other targets around the first target vehicle, the second target having an interaction relationship with the first target vehicle is screened out to construct a target pair, and predictions are made for the targets in each target pair to obtain all possible trajectories of the targets in the target pair. Using the first trajectory and the second trajectory in each target pair to predict the third trajectory can take into account all possible driving situations, making the obtained third trajectory and its occurrence probability more accurate, and adjusting the occurrence probability of the third trajectory according to safety constraint conditions to make the finally obtained optimal trajectory safer and more reliable. And adjusting the optimal trajectory according to driving constraint conditions makes the optimal trajectory more in line with normal human driving habits and improves the comfort of the vehicle driving process.

[0058] Figure 2 It is a schematic diagram of a trajectory planning scenario provided by an embodiment of the present invention.

[0059] As Figure 2 shown in (a), vehicle A is the first target vehicle, and vehicles B and C are other surrounding targets. The first surrounding environment information at the current time is obtained to obtain the map information of the current location of the vehicle, the action information of other targets, and the signal light information. There is no signal light set on the road in this embodiment. Specifically, target B and target C can be identified to determine that target B and target C are cars, and the type labels of target B and target C are added to the first surrounding environment information.

[0060] In a specific implementation, a collision model is used to determine a second target having an interaction relationship with vehicle A based on the first surrounding environment information, and it is determined that vehicle B is the second target, and vehicle C has no interaction relationship with vehicle A. A target pair is formed by vehicle A and vehicle B, and redundant data is removed from the first surrounding environment information and the second surrounding environment information respectively from the main perspectives of vehicle A and vehicle B. The second surrounding environment information refers to the surrounding environment information from the main perspective of vehicle B, and the second surrounding environment information is obtained based on the first surrounding environment information. As Figure 2 shown in (b), for vehicle A, a target pair trajectory prediction model is used to predict at least one first trajectory of the first target vehicle A, and the first trajectories A1 and A2 are obtained; as Figure 2 shown in (c), for vehicle B, a target pair trajectory prediction model is used to predict at least one second trajectory of the second target B, and the second trajectories B1 and B2 are obtained. The trajectories A1, A2, B1, and B2 predicted by the target pair trajectory prediction model may include occurrence probabilities.

[0061] Further, as Figure 2 shown in (d), the trajectories A1, A2, B1, B2 predicted by the target pair trajectory prediction model and the occurrence probabilities of A1, A2, B1, B2 are input into the trajectory prediction model to obtain the third trajectories A3 and A4, and the occurrence probability of A3 is 75% and the occurrence probability of A4 is 25%. The number of the third trajectories is equal to the number of the first trajectories.

[0062] After the trajectory prediction model outputs the third trajectories A3 and A4 and the occurrence probabilities of A3 and A4, the occurrence probabilities of each third trajectory are adjusted according to the safety constraint conditions to reduce the occurrence probabilities of the third trajectories that do not meet the safety constraint conditions. If there is a possibility that the third trajectory A4 collides with an obstacle or the second target B, the occurrence probability of 25% of the third trajectory A4 is adjusted, for example, reduced to 20%. The optimal trajectory is selected according to the occurrence probabilities of the third trajectories A3 and A4, and the A3 with the highest occurrence probability is selected as the optimal trajectory.

[0063] In a specific implementation, after obtaining the optimal trajectory, the acceleration, speed, and turning angle of the first target vehicle A at each moment are calculated according to the optimal trajectory A3. For example, the acceleration of the first target vehicle A from 0 seconds to 1 second is +5 km / h, the turning angle direction is -45°, and the speed is 30 km / h; the acceleration from 1 second to 2 seconds is -10 km / h, the turning angle direction is 0°, and the speed is 20 km / h. The first target vehicle A can determine the specific operations to be performed based on the calculated speed and turning angle.

[0064] Furthermore, after obtaining the optimal trajectory A3, the optimal trajectory A3 can be adjusted according to the driving constraint conditions to limit the speed and turning angle of the first target vehicle A, and control the change frequency of the speed and turning angle of the first target vehicle A within the first preset range. For example, the speed of the first target vehicle A from 0 to 1 second is 30 km / h, the speed from 1 to 2 seconds is 15 km / h, and the speed from 2 to 3 seconds is 40 km / h. At this time, the change in speed is too large, and the speed of the first target vehicle can be adjusted so that the speed from 0 to 3 seconds remains 30 km / h, making the change frequency of the speed within the first preset range, which is more in line with the normal driving habits of humans. The first preset range can be set in advance artificially. Or, the speed of the first target vehicle A from 0 to 1 second is 70 km / h, and the speed is adjusted so that the speed from 0 to 1 second remains 50 km / h, making the optimal trajectory safer.

[0065] Figure 3 It is a schematic diagram of another trajectory planning scenario provided by an embodiment of the present invention.

[0066] As Figure 3 (a) shows that vehicle A is the first target vehicle, and vehicle B and vehicle C are other surrounding targets. The first surrounding environment information at the current time is obtained to obtain the map information of the current location of the vehicle, the action information of other targets, and the signal light information. Specifically, target recognition can be performed on target B and target C to determine that target B is an automobile and target C is a non-motor vehicle, and the type labels of target B and target C are added to the first surrounding environment information.

[0067] In a specific implementation, a collision model is used to determine a second target having an interaction relationship with vehicle A according to the first surrounding environment information, determine that vehicle B and vehicle C are the second targets, form target pairs A-B and A-C between vehicle A and vehicle B and vehicle C respectively, and eliminate redundant data from the first surrounding environment information and the second surrounding environment information.

[0068] As Figure 3 (b) shows that for vehicle A in the A-B target pair, at least one first trajectory of the first target vehicle A is predicted using the target pair trajectory prediction model to obtain the first trajectory A1 and the first trajectory A2; as Figure 3As shown in (c), for vehicle B, at least one second trajectory of the second target B is predicted using the target pair trajectory prediction model, obtaining the second trajectory B1 and the second trajectory B2. The trajectories A1, A2, B1, B2 predicted by the target pair trajectory prediction model may include occurrence probabilities.

[0069] Figure 4 It is a schematic diagram of another trajectory planning scenario provided by an embodiment of the present invention.

[0070] According to Figure 3 the trajectory planning scenario in Figure 4 As shown in (a), for vehicle A in the A-C target pair, at least one first trajectory of the first target vehicle A is predicted using the target pair trajectory prediction model, obtaining the first trajectory A3 and the first trajectory A4; as Figure 4 shown in (b), for vehicle C, at least one second trajectory of the second target C is predicted using the target pair trajectory prediction model, and the target pair trajectory prediction model will refer to the forward direction of the target for the prediction of the second trajectory, obtaining the second trajectory C1 and the second trajectory C2. The trajectories A3, A4, C1, C2 predicted by the target pair trajectory prediction model may include occurrence probabilities.

[0071] Further, the trajectories A1, A2, B1, B2, C1, C2 predicted by the target pair trajectory prediction model and the occurrence probabilities of A1, A2, B1, B2, C1, C2 are input into the trajectory prediction model to obtain the third trajectories A5, A6, A7, A8, and the occurrence probabilities of A5 being 20%, A6 being 30%, A7 being 40%, and A8 being 10%. Specifically, since traffic lights are set on this road, the traffic lights are also used as a reference when predicting the first trajectory, the second trajectory, and the third trajectory to ensure that the first trajectory, the second trajectory, and the third trajectory comply with traffic rules.

[0072] After the trajectory prediction model outputs the third trajectories A5, A6, A7, A8 and the occurrence probabilities of A5, A6, A7, A8, the occurrence probabilities of each third trajectory are adjusted according to the safety constraint conditions to reduce the occurrence probabilities of the third trajectories that do not meet the safety constraint conditions. If there is a possibility that the third trajectory A7 collides with an obstacle, the second target B, or the second target C, the occurrence probability of the third trajectory A7, which is 40%, is adjusted, for example, reduced to 25%. As Figure 4 shown in (c), finally, A6 with the highest occurrence probability is selected as the optimal trajectory.

[0073] In a specific implementation, after obtaining the optimal trajectory A6, the acceleration, speed, and turning angle of the first target vehicle A at each moment are calculated according to the optimal trajectory A6. For example, the acceleration of the first target vehicle A from 0 second to 1 second is -5 km / h, the turning angle direction is 60°, and the speed is 10 km / h; the acceleration from 1 second to 2 seconds is 0 km / h, the turning angle direction is 45°, and the speed is 10 km / h. The first target vehicle A can determine the specific operations to be performed according to the calculated speed and turning angle. For the adjustment of the optimal trajectory A6, reference can be made to Figure 2 the relevant descriptions in

[0074] As Figure 5 shown, an embodiment of the present invention also discloses a trajectory planning device. The trajectory planning device 50 includes:

[0075] An acquisition module 501, configured to acquire the first surrounding environment information of the first target vehicle, where the first surrounding environment information includes the action information of other targets;

[0076] A target pair generation module 502, configured to determine at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information, and form at least one target pair, where each target pair includes the first target vehicle and a second target;

[0077] A first prediction module 503, configured to, for each target pair, predict at least one first trajectory of the first target vehicle according to the first surrounding environment information, and predict at least one second trajectory of the second target according to the second surrounding environment information of the second target;

[0078] A second prediction module 504, configured to predict the third trajectory of the first target vehicle and its occurrence probability according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair, where the occurrence probability represents the probability that the first target vehicle travels according to the third trajectory;

[0079] A selection module 505, configured to select the optimal trajectory of the first target vehicle according to the occurrence probabilities of the respective third trajectories.

[0080] In a specific implementation, the above-mentioned trajectory planning device may correspond to a chip with data processing functions in a server, such as a SOC (System-On-a-Chip), a baseband chip, etc.; or correspond to a chip module including a chip with data processing functions in a server; or correspond to a chip module with a chip having data processing functions, or correspond to a server.

[0081] For more content about the working principle and working mode of the trajectory planning device 50, reference can be made to Figures 1 to 4The relevant descriptions in [reference] are not elaborated here.

[0082] Regarding each device and product described in the above embodiments, each module / unit included therein can be a software module / unit, a hardware module / unit, or can be partially a software module / unit and partially a hardware module / unit. For example, for each device and product applied to or integrated into a chip, each module / unit included therein can be implemented in a hardware manner such as a circuit, or at least some of the modules / units can be implemented in the form of a software program that runs on a processor integrated inside the chip, and the remaining (if any) part of the modules / units can be implemented in a hardware manner such as a circuit; for each device and product applied to or integrated into a chip module, each module / unit included therein can be implemented in a hardware manner such as a circuit, and different modules / units can be located in the same component (such as a chip, a circuit module, etc.) or different components of the chip module, or at least some of the modules / units can be implemented in the form of a software program that runs on a processor integrated inside the chip module, and the remaining (if any) part of the modules / units can be implemented in a hardware manner such as a circuit; for each device and product applied to or integrated into a terminal, each module / unit included therein can be implemented in a hardware manner such as a circuit, and different modules / units can be located in the same component (such as a chip, a circuit module, etc.) or different components inside the terminal, or at least some of the modules / units can be implemented in the form of a software program that runs on a processor integrated inside the terminal, and the remaining (if any) part of the modules / units can be implemented in a hardware manner such as a circuit.

[0083] An embodiment of the present invention also discloses a storage medium. The computer-readable storage medium is a non-volatile storage medium or a non-transitory storage medium. The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program runs, it can execute the steps of the trajectory planning method in the foregoing embodiments. The storage medium can include ROM, RAM, a magnetic disk, an optical disk, etc. The storage medium can also include a non-volatile memory or a non-transitory memory, etc.

[0084] An embodiment of the present invention also discloses a server. The server can include a memory and a processor. A computer program that can run on the processor is stored on the memory. When the processor runs the computer program, it can execute the steps of the trajectory planning method in the foregoing embodiments.

[0085] In the embodiments of the present application, "a plurality of" refers to two or more.

[0086] In the embodiments of the present application, the first, second, etc. descriptions are only used for indicating and distinguishing the described objects, without any order, and do not represent a special limitation on the number of devices in the embodiments of the present application, and cannot constitute any limitation on the embodiments of the present application.

[0087] It should be understood that in the embodiments of the present application, the processor may be a central processing unit (CPU for short), and the processor may also be other general-purpose processors, digital signal processors (DSP for short), application specific integrated circuits (ASIC for short), field programmable gate arrays (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0088] It should also be understood that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM for short), a programmable read-only memory (PROM for short), an erasable programmable read-only memory (EPROM for short), an electrically erasable programmable read-only memory (EEPROM for short), or a flash memory. The volatile memory may be a random access memory (RAM for short), which is used as an external cache. By way of example but not limitation, many forms of random access memory (RAM for short) are available, such as static random access memory (SRAM for short), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM for short), double data rate synchronous dynamic random access memory (DDR SDRAM for short), enhanced synchronous dynamic random access memory (ESDRAM for short), synchlink dynamic random access memory (SLDRAM for short), and direct rambus random access memory (DR RAM for short).

[0089] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired or wireless manner. The computer-readable storage medium can be any available medium that can be accessed by a computer, or a data storage device such as a server or data center that includes one or more collections of available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0090] It should be understood that in various embodiments of the present application, the sequence numbers of the above processes do not imply the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0091] In several embodiments provided in the present application, it should be understood that the disclosed methods, apparatuses, and systems can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings, direct couplings, or communication connections between each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0092] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0093] In addition, in each embodiment of the present invention, each functional unit can be integrated into one processing unit, or each unit can be physically separate, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware, or in the form of a hardware plus software functional unit.

[0094] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit stored in a storage medium includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute some steps of the methods described in each embodiment of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs that can store program codes.

[0095] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention should be subject to the scope defined by the claims.

Claims

1. A trajectory planning method, characterized in that, Including: Obtain the first surrounding environment information of the first target vehicle, where the first surrounding environment information includes the action information of other targets; Determine at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information, to form at least one target pair, and each target pair includes the first target vehicle and a second target; For each target pair, predict at least one first trajectory of the first target vehicle according to the first surrounding environment information, and predict at least one second trajectory of the second target according to the second surrounding environment information of the second target, where the second surrounding environment information refers to the surrounding environment information with the second target as the main perspective, and the second surrounding environment information is obtained based on the first surrounding environment information; Predict the third trajectory and its occurrence probability of the first target vehicle according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair, where the occurrence probability represents the probability that the first target vehicle travels according to the third trajectory; Adjust the occurrence probabilities of the third trajectories according to the safety constraint conditions, so as to reduce the occurrence probabilities of the third trajectories that do not meet the safety constraint conditions; Select the optimal trajectory of the first target vehicle according to the adjusted occurrence probabilities of the third trajectories.

2. The trajectory planning method according to claim 1, wherein The determining at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information includes: Input the first surrounding environment information into a collision model to obtain at least one second target having an interaction relationship with the first target vehicle.

3. The trajectory planning method according to claim 1, wherein The predicting the third trajectory and its occurrence probability of the first target vehicle according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair includes: Input the first trajectory and the second trajectory into a trajectory prediction model to obtain the third trajectory and its occurrence probability of the first target vehicle.

4. The trajectory planning method according to claim 1, wherein The predicting at least one first trajectory of the first target vehicle according to the first surrounding environment information, and predicting at least one second trajectory of the second target according to the second surrounding environment information of the second target includes: Predict the first trajectory using the first surrounding environment information after removing redundant data; Predict the second trajectory using the second surrounding environment information after removing redundant data.

5. The trajectory planning method according to claim 1, characterized in that The obtaining the first surrounding environment information of the first target vehicle includes: Merge the obtained map information, the action information of surrounding targets, and the signal light information to form the first surrounding environment information.

6. The trajectory planning method according to claim 5, wherein The method for obtaining the map information includes: Obtain the map data pre-stored in the database, and determine the map information of the area where the vehicle is located according to the positioning information of the first target vehicle.

7. The trajectory planning method according to claim 1, characterized in that After the selecting the optimal trajectory of the first target vehicle according to the occurrence probabilities of the third trajectories includes: Adjust the optimal trajectory according to the driving constraint conditions, where the driving constraint conditions are used to limit the speed and turning angle of the first target vehicle.

8. The trajectory planning method according to claim 1, wherein Before determining at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information, it includes: Performing target recognition on the other targets to determine the type tags of the other targets; Adding the type tags of the other targets to the first surrounding environment information.

9. A trajectory planning device, characterized by It includes: An acquisition module, configured to acquire first surrounding environment information of a first target vehicle, where the first surrounding environment information includes action information of other targets; A target pair generation module, configured to determine at least one second target having an interaction relationship with the first target vehicle among the other targets according to the first surrounding environment information, and form at least one target pair, where each target pair includes the first target vehicle and a second target; A first prediction module, configured to, for each target pair, predict at least one first trajectory of the first target vehicle according to the first surrounding environment information, and predict at least one second trajectory of the second target according to second surrounding environment information of the second target, where the second surrounding environment information refers to the surrounding environment information from the perspective of the second target, and the second surrounding environment information is obtained based on the first surrounding environment information; A second prediction module, configured to predict a third trajectory of the first target vehicle and its occurrence probability according to the first trajectory of the first target vehicle and the second trajectory of the second target in each target pair, where the occurrence probability represents the probability that the first target vehicle travels according to the third trajectory; A selection module, configured to adjust the occurrence probabilities of the third trajectories according to safety constraint conditions to reduce the occurrence probabilities of the third trajectories that do not meet the safety constraint conditions; The selection module is further configured to select an optimal trajectory of the first target vehicle according to the adjusted occurrence probabilities of the third trajectories.

10. A server, comprising a memory and a processor, wherein a computer program capable of running on the processor is stored on the memory, characterized in that, When the processor runs the computer program, it executes the steps of the trajectory planning method according to any one of claims 1 to 7.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer-readable storage medium is a non-volatile storage medium or a non-transitory storage medium, and when the computer program is run by a processor, it executes the steps of the trajectory planning method according to any one of claims 1 to 7.

Images