A method for processing trajectory queries of a planning module

Abstract

An embodiment of the present invention relates to a method for processing trajectory queries of a planning module. The method includes: the planning module receiving a first set of predicted trajectory groups sent by a prediction module at the start of a first working cycle; generating a set of full-scenario trajectory groups by combining full-obstacle prediction scenarios based on the first set of predicted trajectory groups; generating a set of full-scenario planned trajectories by performing full-scenario vehicle trajectory planning based on the set of full-scenario trajectory groups; receiving a first trajectory query instruction sent by a control module during the first working cycle; locating the scenario trajectory groups in the set of full-scenario trajectory groups that match the first real-time trajectory set and recording them as corresponding first matching scenario trajectory groups; performing effective trajectory interception on the scenario planned trajectories corresponding to the first matching scenario trajectory groups in the set of full-scenario planned trajectories to generate corresponding first query trajectories; and sending the first query trajectories back to the control module as instruction return data. By means of the present invention, the safety guarantee of vehicle driving can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly relates to a method for processing trajectory queries of a planning module. Background Art

[0002] In the conventional working process of the planning module in the automatic driving system of a vehicle, the predicted trajectory of an obstacle is obtained from the upstream prediction module, and the driving trajectory of the vehicle itself is planned based on the predicted trajectory of the obstacle, and then the planned trajectory is output and sent to the downstream control module. The control module then uses the received planned trajectory as a reference trajectory and performs trajectory tracking on it. During the trajectory tracking process, the driving control, that is, the real-time motion state of the vehicle itself is controlled (such as steering wheel control, throttle / brake control, etc.), so that the actual motion trajectory is as close as possible to the reference trajectory. The conventional planning module only outputs the planned trajectory to the control module once within one working cycle.

[0003] In actual applications, we often encounter the situation where the trajectory of an obstacle suddenly changes. The so-called sudden change in the obstacle trajectory means that the actual motion trajectory of the obstacle does not match the predicted trajectory of the obstacle output by the previous prediction module. In this case, the control module should invalidate the current reference trajectory and obtain a new planned trajectory that matches the actual motion trajectory of the obstacle as the reference trajectory, so as to ensure that the driving control instructions it outputs can cope with the current sudden change in the obstacle trajectory. However, the conventional planning module's processing method of single-cycle single-planning output cannot send multiple planned trajectories to the control module within one working cycle. That is to say, according to this conventional processing method, after the control module invalidates the previous reference trajectory, it cannot obtain a new planned trajectory in time as a supplement, which undoubtedly brings great potential safety hazards to the driving control of the vehicle. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for processing trajectory queries of a planning module, an electronic device, and a computer-readable storage medium in view of the defects of the prior art. Through the present invention, the planning module can support the trajectory reverse query operation of the control module, thereby breaking the limitation that the planning module can only output one planned trajectory in a single cycle in the conventional processing method, and improving the safety guarantee for vehicle driving control.

[0005] To achieve the above object, in the first aspect of the embodiments of the present invention, a method for processing trajectory queries of a planning module is provided, and the method includes:

[0006] The planning module receives a first set of predicted trajectory groups sent by the prediction module at the start of the first working cycle;

[0007] Generate a corresponding set of full-scenario trajectory groups by combining all obstacle prediction scenarios according to the first set of predicted trajectory groups;

[0008] Generate a corresponding full-scenario planned trajectory set according to the full-scenario trajectory group set;

[0009] Receive a first trajectory query instruction sent by the control module during the first working cycle; the first trajectory query instruction includes a first real-time trajectory set;

[0010] Locate the scenario trajectory group in the full-scenario trajectory group set that matches the first real-time trajectory set and denote it as the corresponding first matching scenario trajectory group;

[0011] Perform effective trajectory interception on the scenario planned trajectory in the full-scenario planned trajectory set corresponding to the first matching scenario trajectory group to generate a corresponding first query trajectory;

[0012] Send the first query trajectory back to the control module as the instruction return data.

[0013] Preferably, the first predicted trajectory group set includes multiple first predicted trajectory groups P i ; each first predicted trajectory group P i corresponds to an obstacle; the first predicted trajectory group P i includes multiple first predicted trajectories p i,j , each first predicted trajectory p i,j is a possible predicted trajectory of the corresponding obstacle; i is the obstacle index, 1 ≤ i ≤ n, n is the number of obstacles; j is the possible trajectory index, 1 ≤ j;

[0014] The full-scenario trajectory group set includes multiple first scenario trajectory groups F k ; the first scenario trajectory group F k is composed of single first predicted trajectories p i,j of all obstacles; k is the scenario trajectory group index, 1 ≤ k;

[0015] The full-scenario planned trajectory set includes multiple first scenario planned trajectories τ k ; the first scenario planned trajectory τ k corresponds to the first scenario trajectory group F k ; the planned trajectory duration of the first scenario planned trajectory τ k is T th ; the planned trajectory duration T th is much greater than (T p + T c ), T p is the working cycle of the planning module, and T c is the working cycle of the control module;

[0016] The first real-time trajectory set includes multiple first real-time trajectories M i ; each of the first real-time trajectories M i corresponds to an obstacle; the trajectory durations of the first real-time trajectories M i are the same, the start and end times are the same, and the time intervals between adjacent trajectory points are the same.

[0017] Preferably, generating a corresponding full-scenario trajectory set according to the first predicted trajectory set for full obstacle prediction scenarios specifically includes:

[0018] Selecting any one of the first predicted trajectories p i from each of the first predicted trajectory groups P i,j to form a corresponding first scenario trajectory group F k , and determining that the number of the first predicted trajectories p k in any two of the first scenario trajectory groups F i,j is the same and is the number n of the obstacles, and determining that the combination relationships of the first predicted trajectories p k in any two of the first scenario trajectory groups F i,j are different; and forming the full-scenario trajectory set from all the obtained first scenario trajectory groups F k .

[0019] Preferably, generating a corresponding full-scenario planned trajectory set according to the full-scenario trajectory set for full-scenario vehicle trajectory planning specifically includes:

[0020] Traversing each of the first scenario trajectory groups F k of the full-scenario trajectory set, and denoting the currently traversed first scenario trajectory group F k as the current scenario trajectory group, and using each of the first predicted trajectories p i,j in the current scenario trajectory group as an avoidance reference trajectory to plan the traveling trajectory of the host vehicle to obtain a corresponding first scenario planned trajectory τ k ; and forming the full-scenario planned trajectory set from all the obtained first scenario planned trajectories τ k .

[0021] Furthermore, the trajectory point of the first scenario planned trajectory τ k at time t is the host vehicle trajectory point τ k (t); the trajectory point of the first predicted trajectory p i,j at time t is the obstacle trajectory point p i,j (t); the first scenario trajectory group F k corresponding to the first scenario planned trajectory τ kThe combination of trajectory points at time t is the obstacle trajectory point combination F k (t) = [p i＝1,j (t), p i＝2,j (t)…p i＝n,j (t)]; where the ego-vehicle trajectory point τ k (t) and any one of the obstacle trajectory points p k (t) in the corresponding obstacle trajectory point combination F i,j (t) should maintain a trajectory point relative distance above the corresponding safety distance, and the safety distance is constrained by the shapes and relative speeds of the ego-vehicle and the corresponding obstacles. The larger the respective shapes, the larger the safety distance; the faster the relative speed, the larger the safety distance.

[0022] Preferably, positioning the scene trajectory groups in the full-scenario trajectory group set that match the first real-time trajectory set and denoting them as the corresponding first matching scene trajectory groups specifically includes:

[0023] Extracting the start time t i and end time t s of the first real-time trajectory M e in the first real-time trajectory set to form a first time period [t s , t e ;

[0024] Segmenting and extracting the trajectories of each first predicted trajectory p i,j in the full-scenario trajectory group set during the first time period [t s , t e as the corresponding first trajectory segments s i,j ;

[0025] Denoting the first predicted trajectory p i where the first trajectory segment s i,j matching each first real-time trajectory M i,j is located as the first matching predicted trajectory;

[0026] Taking all the first scene trajectory groups F k where the first matching predicted trajectories are located as the scene trajectory groups that match the first real-time trajectory set and denoting them as the corresponding first matching scene trajectory groups.

[0027] Preferably, effectively intercepting the scene planning trajectories corresponding to the first matching scene trajectory groups in the full-scenario planning trajectory set to generate corresponding first query trajectories specifically includes:

[0028] Taking the first scene planning trajectory τ corresponding to the first matching scene trajectory group in the full-scenario planning trajectory setk as the first matching planned trajectory; and intercept all the trajectories in the first matching planned trajectory whose time information is after the current moment as the corresponding first query trajectory.

[0029] Preferably, the planning module can receive and process the trajectory query instruction sent by the control module multiple times within the same working cycle.

[0030] A second aspect of the embodiments of the present invention provides an electronic device, including: a memory, a processor, and a transceiver;

[0031] The processor is used to be coupled with the memory, read and execute the instructions in the memory to implement the method steps described in the first aspect above;

[0032] The transceiver is coupled with the processor, and the processor controls the transceiver to perform message sending and receiving.

[0033] A third aspect of the embodiments of the present invention provides a computer-readable storage medium, which stores computer instructions. When the computer instructions are executed by a computer, the computer is caused to execute the instructions of the method described in the first aspect above.

[0034] An embodiment of the present invention provides a method for processing trajectory queries of a planning module, an electronic device, and a computer-readable storage medium. The planning module receives a first set of predicted trajectory groups sent by a prediction module at the start of a single working cycle. Each first predicted trajectory group in this set corresponds to an obstacle, and each first predicted trajectory group includes predicted trajectories of all possible movements of the corresponding obstacle. The planning module combines all obstacle prediction scenarios based on the first set of predicted trajectory groups to obtain a set of full-scenario trajectory groups. Each first scenario trajectory group in this set corresponds to a full-obstacle movement scenario in a future time period. The planning module performs full-scenario vehicle trajectory planning based on the set of full-scenario trajectory groups to obtain a set of full-scenario planned trajectories. Each first scenario planned trajectory in this set is the driving plan trajectory of the host vehicle in a certain scenario in the future time period. After obtaining the set of full-scenario planned trajectories, the planning module neither performs optimal trajectory screening on the set nor sends the screened trajectories to the control module. Instead, it waits to receive a trajectory query instruction sent by the control module, and the planning module can receive and process the trajectory query instructions sent by the control module multiple times within the same working cycle. Each time it receives a trajectory query instruction, the planning module locates the full-obstacle trajectory group that matches the scenario, i.e., the first matching scenario trajectory group, in the set of full-scenario trajectory groups based on the set of real-time trajectories of all obstacles included in the instruction, i.e., the first real-time trajectory set. Then, based on the corresponding relationship between the scenario trajectory group and the planned trajectory, it locates the planned trajectory that matches the scenario in the set of full-scenario planned trajectories, and intercepts the valid trajectory part of the matching planned trajectory as the instruction return data of the trajectory query instruction and sends it back to the control module. Through the present invention, the planning module realizes the support for the trajectory reverse query operation of the control module, breaks the limitation that the planning module can only output one planned trajectory in a single cycle in the conventional processing method, and improves the safety guarantee for vehicle driving control. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 FIG. is a schematic diagram of a method for processing trajectory queries of a planning module provided in Embodiment 1 of the present invention;

[0036] Figure 2 FIG. is a schematic diagram of the structure of an electronic device provided in Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0037] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below 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. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0038] Embodiment 1 of the present invention provides a method for processing trajectory queries of a planning module, asFigure 1 As shown in the schematic diagram of the trajectory query processing method of a planning module provided in the first embodiment of the present invention, the method mainly includes the following steps:

[0039] Step 1, the planning module receives the first set of predicted trajectory groups sent by the prediction module at the starting moment of the first working cycle;

[0040] Among them, the first set of predicted trajectory groups includes multiple first predicted trajectory groups P i ; each first predicted trajectory group P i corresponds to one obstacle; the first predicted trajectory group P i includes multiple first predicted trajectories p i,j ; each first predicted trajectory p i,j is a predicted trajectory that may occur for the corresponding obstacle; i is the obstacle index, 1 ≤ i ≤ n, where n is the number of obstacles; j is the possible trajectory index, 1 ≤ j.

[0041] Here, let the starting moment of the current working cycle of the planning module be t1. There are multiple obstacles in the driving environment of the vehicle, and the number is n. The duration of the predicted trajectory is agreed to be equal to the duration of the planned trajectory in the subsequent steps as T th , then the first set of predicted trajectory groups is the predicted trajectory results of all possible movements of these n obstacles by the prediction module in the future time period [t1, t1 + T th ; the obstacles here can be any objects, such as buildings, vehicles, traffic devices, road devices, people or animals or plants, non-biological entity objects, etc. In specific implementation, they are classified and defined according to actual needs; the constraint relationship of T th here is that T th is much greater than (T p + T c ), where T p is the working cycle of the planning module, and T c is the working cycle of the control module.

[0042] Each first predicted trajectory group P in the first set of predicted trajectory groups i is the predicted trajectory results of all possible movements of a single obstacle. For example, if obstacle 1 is a vehicle, all its possible future movements may include 4 possibilities: driving left, driving right, going straight, and parking. Then the corresponding first predicted trajectory group P i should include 4 first predicted trajectories p i＝1,j＝1 , p i＝1,j＝2 , p i＝1,j＝3 , p i＝1,j＝4 corresponding to the 4 possible predicted trajectories of driving left, driving right, going straight, and parking respectively; each first predicted trajectory p i,jIt is composed of multiple predicted trajectory points. The time interval between adjacent trajectory points is the same, which is the preset time interval △t. The trajectory information corresponding to each predicted trajectory point should include at least position information and motion information. Among them, the position information can be positioning information such as map coordinates or road coordinates that can identify the position of the obstacle, and the motion information is composed of information that can reflect the motion state of the obstacle, such as speed, acceleration, pose angle, etc.

[0043] Step 2: Generate a corresponding set of full-scenario trajectory groups according to the first set of predicted trajectory groups for the full obstacle prediction scenario combination;

[0044] Among them, the set of full-scenario trajectory groups includes multiple first-scenario trajectory groups F k ; The first-scenario trajectory group F k is composed of a single first predicted trajectory p of all obstacles i,j Combined; k is the scene trajectory group index, 1≤k;

[0045] Specifically, it includes: selecting any one first predicted trajectory p from each first set of predicted trajectory groups P i to form the corresponding first-scenario trajectory group F i,j , and determining that the number of first predicted trajectories p in any two first-scenario trajectory groups F k is the same and is the number of obstacles n, and determining that the combination relationship of the first predicted trajectories p in any two first-scenario trajectory groups F k is different; and forming a set of full-scenario trajectory groups from all the obtained first-scenario trajectory groups F i,j k i,j k th th k .

[0046] Here, in fact, it is to combine all possible motions of all obstacles in the future time period [t1, t1 + T th according to the first set of predicted trajectory groups, so that each combination relationship can represent a full obstacle motion scenario in the future time period [t1, t1 + T th . Each first-scenario trajectory group F k is the trajectory collection of all obstacles in a single scenario, and the set of full-scenario trajectory groups composed of all first-scenario trajectory groups F k is the trajectory collection of all obstacles in all scenarios. It should be noted that the total number of combinations of the set of full-scenario trajectory groups is k max , and k max = max(j i＝1 )×max(j i＝2 )…×max(j i＝n ); Among them, max(j i＝1 ) represents the first predicted trajectory p corresponding to when i = 1 i＝1,jThe maximum value of the subscript j is the total number of possible types of movement of obstacle 1, max(j i＝2 ) represents the first predicted trajectory p corresponding to when i = 2 i＝2,j The maximum value of the subscript j is the total number of possible types of movement of obstacle 2, and so on, max(j i＝n ) represents the first predicted trajectory p corresponding to when i = the number of obstacles n i＝n,j The maximum value of the subscript j is the total number of possible types of movement of obstacle n.

[0047] For example, there are n = 2 obstacles 1 and 2 in the self-vehicle driving environment, and the first predicted trajectory group set includes the first predicted trajectory groups P1 and P2; obstacle 1 has 2 possible movements (left, right) in the future time period, then the first predicted trajectory group P1 includes 2 first predicted trajectories p 1,1 、p 1,2 ; if obstacle 1 has 3 possible movements (left, right, straight) in the future time period, then the first predicted trajectory group P2 includes 3 first predicted trajectories p 2,1 、p 2,2 、p 2,3 ;

[0048] Then, the total combination quantity k of the full-scenario trajectory group set max = max(j i＝1 ) × max(j i＝2 ) = 2 × 3 = 6, and the full-scenario trajectory group set includes 6 first-scenario trajectory groups F1, F2, F3, F4, F5, F6, respectively representing 6 combined scenarios of 2 obstacles within the future time period [t1, t1+T th :

[0049] F1 = (p 1,1 , p 2,1 ), Scenario 1, within the future time period [t1, t1+T th , obstacle 1 moves left and obstacle 2 moves left;

[0050] F2 = (p 1,1 , p 2,2 ), Scenario 2, within the future time period [t1, t1+T th , obstacle 1 moves left and obstacle 2 moves right;

[0051] F3 = (p 1,1 , p 2,3 ), Scenario 3, within the future time period [t1, t1+T th , obstacle 1 moves left and obstacle 2 moves straight;

[0052] F4 = (p 1,2 , p 2,1), Scenario Four, in the future time period [t1, t1 + T th , obstacle 1 moves to the right and obstacle 2 moves to the left;

[0053] F5 = (p 1,2 , p 2,2 ), Scenario Five, in the future time period [t1, t1 + T th , obstacle 1 moves to the right and obstacle 2 moves to the right;

[0054] F6 = (p 1,2 , p 2,3 ), Scenario Six, in the future time period [t1, t1 + T th , obstacle 1 moves to the right and obstacle 2 moves straight ahead.

[0055] Step 3, perform full-scenario vehicle trajectory planning based on the full-scenario trajectory group set to generate the corresponding full-scenario planned trajectory set;

[0056] Among them, the full-scenario planned trajectory set includes multiple first-scenario planned trajectories τ k ; The first-scenario planned trajectory τ k corresponds to the first-scenario trajectory group F k ; The planned trajectory duration of the first-scenario planned trajectory τ k is T th ; The planned trajectory duration T th is much greater than (T p + T c ), T p is the working cycle of the planning module, and T c is the working cycle of the control module;

[0057] Specifically, it includes: traversing each first-scenario trajectory group F k in the full-scenario trajectory group set, and recording the currently traversed first-scenario trajectory group F k as the current scenario trajectory group, and using each first predicted trajectory p i,j in the current scenario trajectory group as the avoidance reference trajectory to plan the traveling trajectory of the host vehicle to obtain the corresponding first-scenario planned trajectory τ k ; And all the first-scenario planned trajectories τ k obtained form the full-scenario planned trajectory set;

[0058] Among them, the trajectory point of the first-scenario planned trajectory τ k at time t is the host vehicle trajectory point τ k (t); The trajectory point of the first predicted trajectory p i,j at time t is the obstacle trajectory point p i,j (t); The first-scenario trajectory group F corresponding to the first-scenario planned trajectory τ k ​k The combination of trajectory points at time t is the obstacle trajectory point combination F k (t)=[p i＝1,j (t), p i＝2,j (t)…p i＝n,j (t)]; where the ego-vehicle trajectory point τ k (t) and any obstacle trajectory point p k (t) in the corresponding obstacle trajectory point combination F i,j (t) should maintain a trajectory point relative distance above the corresponding safety distance, and the safety distance is constrained by the shapes and relative speeds of the ego-vehicle and the corresponding obstacles. The larger the respective shapes, the larger the safety distance; the faster the relative speed, the larger the safety distance.

[0059] Here, since the full-scenario trajectory group set includes all possible predicted trajectories of all actions of all obstacles in the future time period, each first-scenario trajectory group F k corresponds to a full obstacle motion scenario in the future. Then, based on each first-scenario trajectory group F k the first-scenario planned trajectory τ k planned is also the ego-vehicle trajectory planning result in the future time period under the corresponding scenario. The full-scenario planned trajectory set composed of all first-scenario planned trajectories τ k can include all possible ego-vehicle planned trajectories in all possible scenarios in the future time period. When planning the ego-vehicle's traveling trajectory using each first predicted trajectory p i,j in the current scenario trajectory group as the avoidance reference trajectory, the planning algorithm used is similar to the trajectory planning algorithm used by the conventional planning module. That is, after obtaining the start and end positions of the current planning, taking each trajectory point of the avoidance reference trajectory as the avoidance target, and combining the driving state of the ego-vehicle to perform the shortest path and motion state planning to obtain the corresponding planned trajectory. The specific implementation can refer to the traditional planning algorithm technology implementation and will not be elaborated further here.

[0060] In addition, the embodiment of the present invention takes into account the safety distance between the ego-vehicle and the obstacle, and when planning, it makes a conditional constraint on the relative distance between the trajectory points of the first-scenario planned trajectory τ k and the corresponding equal-time trajectory points on each first predicted trajectory p k in the first-scenario trajectory group F i,j . The constraint condition is that the trajectory point p k at any time t of the first-scenario planned trajectory τ i,j (t) and each first predicted trajectory p k in the corresponding first-scenario trajectory group F i,j at the corresponding equal-time trajectory point p i,jThe relative distance between (t) should not be less than a predefined safe distance. Since the distance between trajectory points is actually the distance between the centers of mass of the objects, this safe distance is actually the safe distance between the ego vehicle and the obstacle's center of mass. Therefore, this safe distance is not only constrained by the relative speed between the ego vehicle and the obstacle but also by the shapes of the ego vehicle and the obstacle. The larger the shapes of the ego vehicle and the obstacle respectively, the larger this safe distance; the faster the relative speed, the larger this safe distance.

[0061] It should be noted that the embodiments of the present invention can implement the constraint on the safe distance based on the shapes and relative speeds of the ego vehicle and the obstacle in various ways; one implementation method is as follows. Constraining the safe distance based on the shapes and relative speeds of the ego vehicle and the obstacle specifically includes: presetting a minimum relative speed and a minimum buffer distance; denoting the longest parameter among the shape parameters (length, width, height, diameter, etc.) of the ego vehicle and the current obstacle as the corresponding first and second parameters, and denoting the relative speed between the ego vehicle and the current obstacle as the corresponding first speed; and calculating a spacing parameter L according to the first and second parameters and the minimum buffer distance, where L = α * (first parameter + second parameter + minimum buffer distance), and α is a preset linear coefficient, α > 0; and when the first speed is less than the minimum relative speed, constraining the safe distance between the ego vehicle and the obstacle to be not less than the spacing parameter L; and when the first speed is greater than or equal to the minimum relative speed, constraining the safe distance to be not less than β * L, where β is a preset linear parameter, β > 1.

[0062] After the planning module of the embodiment of the present invention finishes processing the above steps 1 - 3, it does not continue to screen out the optimal planned trajectory from the full - scene planned trajectory set and output it to the control module, but waits to receive and process the trajectory query instruction sent by the control module.

[0063] Step 4, receiving the first trajectory query instruction sent by the control module during the first working cycle;

[0064] Among them, the first trajectory query instruction includes a first real - time trajectory set; the first real - time trajectory set includes multiple first real - time trajectories M i ; each first real - time trajectory M i corresponds to an obstacle; the trajectory durations of each of the first real - time trajectories M i are the same, the start and end times are the same, and the time intervals between adjacent trajectory points are the same.

[0065] Here, the first real - time trajectory set is a section of the real - time motion trajectory of all obstacles obtained by the control module from the vehicle's perception sensor or perception module. Let the time when the control module sends the first trajectory query instruction be time t2, and this time t2 should be later than the start time t1 of the current working cycle of the planning module mentioned above; let the start time of each of the first real - time trajectories M i be ts and the end time is t e , so t s should not be later than t e ; also, since the first real-time trajectory set is a trajectory segment where all obstacles have occurred, so t e should not be later than t2; also, since the subsequent steps need to use the first real-time trajectory set to query for matching trajectories in the full-scenario trajectory group set, so t s cannot be earlier than t1; in summary, the correlation relationship of all the above time parameters should be as follows: t1 ≤ t s ≤ t e ≤ t2; in the specific implementation process, usually the above time parameters are set as follows: t1 < t s < t e ≤ t2, and the end time t e should be as close to t2 as possible, so as to ensure the real-time nature of the first real-time trajectory set.

[0066] Step 5, locate the scenario trajectory groups in the full-scenario trajectory group set that match the first real-time trajectory set and denote them as the corresponding first matching scenario trajectory groups;

[0067] Specifically, it includes: Step 51, extract the start time t i and the end time t s of the first real-time trajectory M e in the first real-time trajectory set to form the first time period [t s , t e ;

[0068] For example, if there are n = 2 obstacles 1 and 2 in the driving environment of the vehicle itself, then the first real-time trajectory set includes 2 first real-time trajectories M1 and M2;

[0069] Step 52, extract the trajectory segments of each first predicted trajectory p i,j in the first real-time trajectory set within the first time period [t s , t e as the corresponding first trajectory segments s i,j ;

[0070] For example, the full-scenario trajectory group set includes 6 first scenario trajectory groups:

[0071] F1 = (p 1,1 , p 2,1 ), F2 = (p 1,1 , p 2,2 ), F3 = (p 1,1 , p 2,3 ),

[0072] F4 = (p 1,2 , p2,1 ), F5 = (p 1,2 , p 2,2 ), F6 = (p 1,2 , p 2,3 );

[0073] Among them, there are a total of 5 first predicted trajectories p 1,1 , p 1,2 , p 2,1 , p 2,2 , p 2,3 ,

[0074] For these 5 first predicted trajectories, extracting the trajectory segments in the first time period [t s , t e will obtain 5 first trajectory segments s 1,1 , s 1,2 , s 2,1 , s 2,2 , s 2,3 ;

[0075] Step 53, mark the first predicted trajectory p i where the first trajectory segment s i,j that matches each first real-time trajectory M i,j as the first matching predicted trajectory;

[0076] For example, the first real-time trajectory set has 2 first real-time trajectories M1 and M2; use the first real-time trajectory M1 to compare with the 2 first trajectory segments s 1,1 , s 1,2 of obstacle 1 one by one. If the matching trajectory segment is the first trajectory segment s 1,1 , then the corresponding first matching predicted trajectory 1 is the first predicted trajectory p 1,1 where s 1,1 is located; use the first real-time trajectory M2 to compare with the 3 first trajectory segments s 2,1 , s 2,2 , s 2,3 of obstacle 2 one by one. If the matching trajectory segment is the first trajectory segment s 2,2 , then the corresponding first matching predicted trajectory 2 is the first predicted trajectory p 2,2 where s 2,2 is located;

[0077] Step 54, regard the first scene trajectory group F k where all the first matching predicted trajectories are located as the scene trajectory group that matches the first real-time trajectory set and mark it as the corresponding first matching scene trajectory group.

[0078] For example, the first matching predicted trajectory 1 is the first predicted trajectory p 1,1, the first matching predicted trajectory 2 is the first predicted trajectory p 2,2 ; then, the first scenario trajectory group F to which all the first matching predicted trajectories belong k should be F2 = (p 1,1 , p 2,2 ), that is, the first matching scenario trajectory group is F2.

[0079] Step 6, perform effective trajectory truncation on the scenario planning trajectories corresponding to the first matching scenario trajectory group in the full-scenario planning trajectory set to generate corresponding first query trajectories;

[0080] Specifically, it includes: taking the first scenario planning trajectory τ corresponding to the first matching scenario trajectory group in the full-scenario planning trajectory set k as the first matching planned trajectory; and truncating all the trajectories in the first matching planned trajectory whose time information is after the current moment as the corresponding first query trajectory.

[0081] Here, because the first scenario planning trajectory τ in the full-scenario planning trajectory set k corresponds one-to-one with the first scenario trajectory group F in the full-scenario trajectory group set k , then after obtaining the first scenario trajectory group F k , the corresponding first scenario planning trajectory τ can naturally be located from the full-scenario planning trajectory set according to the subscript k k . Also, because the planned trajectory output to the control module should be the trajectory in the future time period, so the obtained first scenario planning trajectory τ k needs to be truncated by time period.

[0082] For example, if the first matching scenario trajectory group is F2, then the first matching planned trajectory should be the first scenario planning trajectory τ2; the full time period of the first scenario planning trajectory τ2 should be the time period [t1, t1 + T th , the time when the control module sends the first trajectory query instruction is the moment t2, and t1 < t2 < t1 + T th ; the current moment t ’ is after the control module sends the first trajectory query instruction, so t2 ≤ t ’ < t1 + T th ; the first query trajectory is the trajectory segment of the first scenario planning trajectory τ2 from t ’ backward to t1 + T th .

[0083] Step 7, send the first query trajectory back to the control module as the instruction return data.

[0084] Here, after receiving the first query trajectory, the control module will use it as the latest reference trajectory and perform driving control on the vehicle based on this.

[0085] In the embodiment of the present invention, the planning module only needs to execute the above steps 1-3 once within the same working cycle; however, it can receive and process the trajectory query instructions sent by the control module multiple times, and each time a new trajectory query instruction is received, the above steps 4-7 are repeatedly executed correspondingly.

[0086] Figure 2 It is a schematic structural diagram of an electronic device provided in the second embodiment of the present invention. The electronic device can be the aforementioned terminal device or server, or can be a terminal device or server connected to the aforementioned terminal device or server to implement the method embodiment of the present invention. As Figure 2 shown, the electronic device may include: a processor 301 (such as a CPU), a memory 302, and a transceiver 303; the transceiver 303 is coupled to the processor 301, and the processor 301 controls the transceiver operations of the transceiver 303. Various instructions can be stored in the memory 302 for completing various processing functions and implementing the processing steps described in the foregoing method embodiments. Preferably, the electronic device related to the embodiment of the present invention further includes: a power supply 304, a system bus 305, and a communication port 306. The system bus 305 is used to implement communication connections between components. The above communication port 306 is used for the electronic device to communicate and connect with other peripherals.

[0087] In Figure 2 the system bus 305 mentioned can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This system bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 2 only a thick line is used to represent it in

[0088] The above-mentioned processor may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), a Graphics Processing Unit (GPU), etc.; it may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0089] It should be noted that the embodiment of the present invention also provides a computer-readable storage medium, in which instructions are stored. When it runs on a computer, it enables the computer to execute the methods and processing procedures provided in the above embodiments.

[0090] The embodiment of the present invention also provides a chip for running instructions, and this chip is used to execute the processing steps described in the foregoing method embodiments.

[0091] The embodiment of the present invention provides a method for processing trajectory query of a planning module, an electronic device, and a computer-readable storage medium. In the embodiment of the present invention, the planning module completes the combination of full-obstacle prediction scenarios according to the first predicted trajectory group set sent by the prediction module at the start moment of a single working cycle to obtain a full-scenario trajectory group set, and completes the full-scenario vehicle trajectory planning based on the full-scenario trajectory group set to obtain a full-scenario planned trajectory set; and after obtaining the full-scenario trajectory group set and the full-scenario planned trajectory set, based on the first real-time trajectory set in the trajectory query instruction sent by the control module, select the predicted scenario trajectory group that best matches the current real-time trajectory from the full-scenario trajectory group set, obtain the planned trajectory corresponding to the matching scenario from the full-scenario planned trajectory set, and use the trajectory of the matching planned trajectory after the current moment as the valid trajectory to send back to the control module. Through the present invention, the limitation that the planning module can only output a planned trajectory once in a single cycle in the conventional processing method is broken, and the planning module can respond to the trajectory reverse query operation of the control module multiple times within the same working cycle, thereby ensuring that the control module will not have the problem of being unable to update the planned trajectory in time, and improving the safety guarantee for vehicle driving control.

[0092] Those skilled in the art should further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0093] The steps of the methods or algorithms described in combination with the embodiments disclosed herein can be implemented by hardware, software modules executed by a processor, or a combination of the two. The software modules can be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0094] The specific embodiments described above have further elaborated on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for processing trajectory queries of a planning module, characterized in that The method includes: The planning module receives a first set of predicted trajectory groups sent by the prediction module at the start of the first working cycle; Performs full-obstacle prediction scenario combination based on the first set of predicted trajectory groups to generate a corresponding set of full-scenario trajectory groups; Performs full-scenario vehicle trajectory planning based on the set of full-scenario trajectory groups to generate a corresponding set of full-scenario planned trajectories; Receives a first trajectory query instruction sent by the control module during the first working cycle; the first trajectory query instruction includes a first set of real-time trajectories; Locates the scenario trajectory groups in the set of full-scenario trajectory groups that match the first set of real-time trajectories and designates them as corresponding first matching scenario trajectory groups; Performs valid trajectory truncation on the scenario planned trajectories in the set of full-scenario planned trajectories that correspond to the first matching scenario trajectory groups to generate corresponding first query trajectories; Returns the first query trajectories as instruction return data and sends them back to the control module; Among them, the first predicted trajectory group set includes a plurality of first predicted trajectory groups P i ; each of the first predicted trajectory groups P i corresponds to an obstacle; the first predicted trajectory group P i includes a plurality of first predicted trajectories p i,j , each of the first predicted trajectories p i,j is a possible predicted trajectory corresponding to the obstacle; i is the obstacle index, 1 ≤ i ≤ n, where n is the number of obstacles; j is the possible trajectory index, 1 ≤ j; The full-scenario trajectory group set includes a plurality of first-scenario trajectory groups F k ; The first-scenario trajectory group F k is composed of a single first prediction trajectory p of all obstacles i,j combined; k is the scenario trajectory group index, 1 ≤ k; The full-scenario planned trajectory set includes multiple first-scenario planned trajectories τ k ; the first-scenario planned trajectory τ k corresponds to the first-scenario trajectory group F k ; the planned trajectory duration of the first-scenario planned trajectory τ k is T th ; the planned trajectory duration T th is much greater than (T p + T c ), where T p is the working cycle of the planning module, and T c is the working cycle of the control module; The first real-time trajectory set includes a plurality of first real-time trajectories M i ; each of the first real-time trajectories M i corresponds to an obstacle; the trajectory durations of the respective first real-time trajectories M i are the same, the start and end times are the same, and the time intervals between adjacent trajectory points are the same; The performing valid trajectory truncation on the scenario planned trajectories in the set of full-scenario planned trajectories that correspond to the first matching scenario trajectory groups to generate corresponding first query trajectories specifically includes: Take the first scenario planning trajectory τ corresponding to the first matching scenario trajectory group in the full-scenario planning trajectory set k as the first matching planning trajectory; and extract all the trajectories in the first matching planning trajectory whose time information is after the current moment as the corresponding first query trajectory.

2. The trajectory query processing method of the planning module according to claim 1, wherein The performing full-obstacle prediction scenario combination based on the first set of predicted trajectory groups to generate a corresponding set of full-scenario trajectory groups specifically includes: Select any one of the first predicted trajectory groups P i to form the corresponding first scenario trajectory group F i,j and determine that the number of the first predicted trajectories p k in any two of the first scenario trajectory groups F k is the same and equal to the number of obstacles n, and determine that the combination relationships of the first predicted trajectories p i,j in any two of the first scenario trajectory groups F k are different; and form the set of all scenario trajectory groups from all the obtained first scenario trajectory groups F i,j k to form the set of all scenario trajectory groups.​ 3. The trajectory query processing method of the planning module according to claim 1, wherein The performing full-scenario vehicle trajectory planning based on the set of full-scenario trajectory groups to generate a corresponding set of full-scenario planned trajectories specifically includes: For each of the first scenario trajectory groups F in the full-scenario trajectory group set k perform a traversal, and denote the currently traversed first scenario trajectory group F k as the current scenario trajectory group, and use each of the first prediction trajectories p in the current scenario trajectory group i,j as an avoidance reference trajectory to plan the driving trajectory of the host vehicle to obtain the corresponding first scenario planned trajectory τ k ; and all the obtained first scenario planned trajectories τ k constitute the full-scenario planned trajectory set.

4. The trajectory query processing method of the planning module according to claim 3, wherein The first scenario planning trajectory τ k The trajectory point at time t is the ego vehicle trajectory point τ k (t); The first predicted trajectory p i,j The trajectory point at time t is the obstacle trajectory point p i,j (t); The first scenario planning trajectory τ k The corresponding first scenario trajectory group F k The trajectory point combination at time t is the obstacle trajectory point combination F k (t) = [p i＝1,j (t), p i＝2,j (t) … p i＝n,j (t)]; Wherein, the ego vehicle trajectory point τ k (t) and any obstacle trajectory point p k (t) in the corresponding obstacle trajectory point combination F i,j (t) should maintain a trajectory point relative distance above the corresponding safety distance, and the safety distance is constrained by the shapes and relative speeds of the ego vehicle and the corresponding obstacle. The larger the respective shapes, the larger the safety distance, and the faster the relative speed, the larger the safety distance.

5. The trajectory query processing method of the planning module according to claim 1, characterized in that The locating the scenario trajectory groups in the set of full-scenario trajectory groups that match the first set of real-time trajectories and designating them as corresponding first matching scenario trajectory groups specifically includes: Extract the first real-time trajectory M in the first real-time trajectory set i of the starting time t s , the ending time t e to form the first time period [t s , t e ; Extract each of the first predicted trajectories p in the full-scenario trajectory group set i,j in the first time period [t s , t e as the corresponding first trajectory segment s i,j ; The first real-time trajectory M corresponding to each one i The first trajectory segment s that matches i,j The first predicted trajectory p where the segment is located i,j is denoted as the first matching predicted trajectory; Regarding all the first scenario trajectory groups F where the first matching predicted trajectories are located k as the scenario trajectory groups matching the first real-time trajectory set and denoted as the corresponding first matching scenario trajectory groups.

6. The trajectory query processing method of the planning module according to claim 1, wherein The planning module can receive and process the trajectory query instructions sent by the control module multiple times within the same working cycle.

7. An electronic device, characterized in that, Includes: A memory, a processor, and a transceiver; The processor is used to be coupled with the memory, read and execute the instructions in the memory to implement the method according to any one of claims 1-6; The transceiver is coupled with the processor, and the processor controls the transceiver to perform message sending and receiving.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a computer, the computer is caused to execute the method according to any one of claims 1-6.

Images