A processing method for trajectory planning based on multi-target predicted trajectories

By receiving and processing a multi-objective predicted trajectory set in the vehicle autonomous driving system, confirming and saving all possible scene trajectory combinations and optimal driving trajectory, the problem that traditional systems cannot cope with the target trajectory mutation is solved, and the safety of the vehicle is improved. Trajectory query capability.

CN114802302BActive Publication Date: 2024-11-15SUZHOU QINGZHOU ZHIHANG INTELLIGENT TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210436549.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-25
Publication Date
2024-11-15
Estimated Expiration
2042-04-25

AI Technical Summary

Technical Problem

When traditional vehicle autonomous driving systems handle multi-target prediction trajectory, they cannot effectively deal with the sudden change in the target trajectory, resulting in safety hazards in driving control. Only one planned trajectory can be settled to meet multiple possible motion scenarios, affecting safety.

Method used

The planning module receives the multi-objective prediction trajectory set output by the upstream prediction module, confirms the combination of all possible scene trajectories in the future period, calculates the optimal driving trajectory in each scenario, and performs trajectory rationality checks to save all optimal planned trajectory sets. , so as to enable real-time query and trajectory switching.

Benefits of technology

This avoids the problem of only one planning trajectory being deposited in the traditional method, provides a complete trajectory query basis, improves the safety guarantee of the vehicle, and ensures that the optimal trajectory planning and switching can be achieved in multiple scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114802302B_ABST
    Figure CN114802302B_ABST
Patent Text Reader

Abstract

The embodiment of the present invention relates to a processing method for trajectory planning based on multi-target predicted trajectories, the method comprising: a planning module at any time t 0 Receive the multi-target prediction trajectory set output by the upstream prediction module; combine any possible prediction trajectories of all targets in the future period T and regard each combination as a possible future scene to obtain scene trajectory combinations f corresponding to multiple future scenes x ; Plan the optimal driving trajectory of the vehicle in the corresponding scene x in the future time period T; Check the rationality of the trajectory; If the check is successful, save the planned trajectory set. Through the present invention, the optimal planned trajectory in any possible scenario in the future time period can be saved as the trajectory data to be checked, which can not only avoid the problem of only depositing one planned trajectory in the conventional processing method, but also provide a complete query basis for real-time trajectory query, thereby achieving the purpose of improving vehicle safety.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a processing method for trajectory planning based on multi-target predicted trajectories. Background Art

[0002] The upstream and downstream modules of the planning module of the vehicle automatic driving system are: prediction module and control module; the traditional prediction module predicts multiple possible motion trajectories of each target in the future period based on the target (obstacle) trajectory information collected by the perception sensor, and sends the predicted multi-target prediction trajectory set to the planning module; the traditional planning module estimates the maximum probability of various possible multi-target motion scenarios in the future period based on the multi-target prediction trajectory set, and outputs the unique trajectory planned based on the maximum probability scenario to the downstream control module; the traditional control module uses the unique trajectory as the reference trajectory and tracks it. During the trajectory tracking process, the actual motion trajectory is made as close to the reference trajectory as possible through driving control, that is, controlling the real-time motion state of the vehicle (such as steering wheel control, throttle / brake control, etc.). In this traditional processing method, the vehicle automatic driving system only deposits one planning trajectory in one working cycle. If the target trajectory suddenly changes, that is, the actual motion scene of the multi-target does not match the estimated maximum probability scene, it will not be possible to query and switch the planning trajectory based on the obtained multi-target actual motion trajectory set. This undoubtedly brings great safety hazards to the vehicle's driving control. Summary of the invention

[0003] The purpose of the present invention is to provide a processing method, electronic device and computer-readable storage medium for trajectory planning based on multi-target predicted trajectories in view of the defects of the prior art. The planning module first confirms the scene trajectory combination of all possible multi-target motion scenes in the future time period based on the multi-target predicted trajectory set sent by the upstream prediction module, and then plans the optimal driving trajectory of the vehicle in each scene in the future time period according to each scene trajectory combination, and then performs a trajectory rationality check on the obtained multiple optimal planned trajectories, and after the check is successful, all the optimal planned trajectories are combined into a corresponding planned trajectory set and saved. Through the present invention, the optimal planned trajectory in any possible scenario in the future time period can be saved as the trajectory data to be checked, which not only avoids the problem of only precipitating one planned trajectory in the conventional processing method, but also provides a complete query basis for real-time trajectory query, thereby achieving the purpose of improving vehicle safety.

[0004] To achieve the above object, a first aspect of an embodiment of the present invention provides a processing method for trajectory planning based on multi-target predicted trajectories, the method comprising:

[0005] The planning module receives the multi-target prediction trajectory set output by the upstream prediction module at any time t0; the prediction trajectory period of the multi-target prediction trajectory set is the future period T: [t0, t0+L], t0 is the start time of the period, t0+L is the end time of the period, and L is the duration of the period; the multi-target prediction trajectory set includes multiple first target trajectory groups R; the first target trajectory group R includes multiple first target trajectory p i,j , i is the target index, j is the trajectory index of target i, 1≤i≤n, 1≤j≤m i , n is the number of targets, m i is the number of trajectories of target i;

[0006] According to the multi-target predicted trajectory set, any possible predicted trajectory of all targets in the future period T is combined and each combination is regarded as a possible future scene, so as to obtain scene trajectory combinations f corresponding to multiple future scenes respectively. x ; x is the scene index, 1≤x≤w, w is the number of scenes; the scene trajectory combination f x The first target trajectory p of the target number n i,j The first target trajectory p in the combination i,j The target index i is not repeated;

[0007] According to each of the scene trajectory combinations f x Plan the optimal driving trajectory of the vehicle in the corresponding scene x in the future period T to obtain the corresponding optimal planning trajectory d x ;

[0008] For all the optimal planning trajectories d x Conduct trajectory plausibility checks;

[0009] If the trajectory rationality check succeeds, then all the optimal planned trajectories d x The planned trajectory set composed of the above two components is saved.

[0010] Preferably, each of the first target trajectories p i,j is a possible motion trajectory of the corresponding target i in the future period T; any two of the first target trajectories p in each of the first target trajectory groups R i,j There is an overlapping trajectory between them starting from the starting position.

[0011] Preferably, according to the multi-target predicted trajectory set, any possible predicted trajectories of all targets in the future period T are combined and each combination is regarded as a possible future scene, so as to obtain scene trajectory combinations f corresponding to multiple future scenes respectively. x , specifically including:

[0012] Select one of the first target trajectories p from each of the first target trajectory groups R of the multi-target trajectory set i,j Combine to form the corresponding scene trajectory combination f x ;

[0013] Among them, each of the scene trajectory combinations f x Each of them corresponds to a possible future scene x in the future period T; any two scene trajectory combinations f x The first target trajectory p in each combination i,j The combination relationship is not repeated; the number of scenes is a combination function.

[0014] Preferably, the combination of each of the scene trajectories f x Plan the optimal driving trajectory of the vehicle in the corresponding scene x in the future period T to obtain the corresponding optimal planning trajectory d x , specifically including:

[0015] Obtaining the starting reference position and the target reference position of the vehicle in the future time period T;

[0016] In the case of not meeting the current scene trajectory combination f x Any of the first target trajectories p i,j Under the condition of a collision, the driving trajectory of the vehicle from the starting reference position to the target reference position in the future period T is planned to obtain one or more planned trajectories d x,a ; a is the planning trajectory index of scene x, 1≤a;

[0017] For the planned trajectory d x,a Counting the number of , generates a corresponding first number;

[0018] Identify the first number; if the first number is 1, then the unique planned trajectory d x,a As the optimal planning trajectory d corresponding to the current scene x x Output; if the first number is greater than 1, then calculate the cost of each planned trajectory d according to the preset trajectory cost calculation principle x,a The trajectory cost is calculated to generate the corresponding first cost data, and the planned trajectory d corresponding to the first cost data with the smallest value is x,a As the optimal planning trajectory d corresponding to the current scene x x Output.

[0019] Preferably, for all the optimal planning trajectories d x Perform trajectory rationality checks, including:

[0020] For all the optimal planning trajectories d x Perform trajectory motion rationality check; if the trajectory motion direction check is successful, then for any two of the optimal planning trajectories d x Perform a trajectory branch time rationality check; if the trajectory branch time check is successful, then the trajectory rationality check is successful.

[0021] Furthermore, for all the optimal planning trajectories d x Perform trajectory motion rationality check, including:

[0022] In each of the optimal planning trajectories d x The previous trajectory point is recorded as the previous trajectory point, and the next trajectory point is recorded as the next trajectory point. The relative displacement of the next trajectory point relative to the previous trajectory point is calculated. If the current optimal planning trajectory d x If all the relative displacements are greater than 0, the corresponding inspection state is set to the pass state; if all the optimal planning trajectories d x If all the corresponding inspection states are passed, the trajectory motion rationality check is successful.

[0023] Furthermore, for any two of the optimal planning trajectories d x Perform a reasonableness check on the trajectory branch time, including:

[0024] Choose any two of the optimal planning trajectories d x The scene trajectories corresponding to the first and second trajectories are combined into f x They are recorded as the corresponding first and second combinations respectively;

[0025] In the first and second combinations, the two first target trajectories p corresponding to each target i i,j Confirm the overlapped trajectory starting from the starting position and use the confirmed overlapped trajectory end time as the corresponding first target trajectory branch time t 1,i , so that n branch times t of the first target trajectory can be obtained 1,i ; and select the earliest first target trajectory branch time t 1,i as a first branch time t1 corresponding to the first and second combinations;

[0026] Confirming the overlapping trajectory of the first and second trajectories starting from the starting position and using the confirmed ending time of the overlapping trajectory as the corresponding second branching time t2;

[0027] Calculate the time difference △t=t2-t1 between the second branch time t2 and the first branch time t1; and identify whether the time difference △t satisfies the preset reasonable time difference range. If the time difference △t satisfies the reasonable time difference range, the trajectory branch time rationality check is successful; the reasonable time difference range is a time difference range greater than 0 and less than or equal to the specified threshold δ.

[0028] A second aspect of an embodiment of the present invention provides an electronic device, including: a memory, a processor, and a transceiver;

[0029] The processor is used to be coupled to the memory, read and execute instructions in the memory, so as to implement the method steps described in the first aspect above;

[0030] The transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.

[0031] A third aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions. When the computer instructions are executed by a computer, the computer executes the instructions of the method described in the first aspect above.

[0032] The embodiment of the present invention provides a processing method, electronic device and computer-readable storage medium for trajectory planning based on multi-target predicted trajectories. The planning module first confirms the scene trajectory combination of all possible multi-target motion scenes in the future time period based on the multi-target predicted trajectory set sent by the upstream prediction module, and then plans the optimal driving trajectory of the vehicle in each scene in the future time period according to each scene trajectory combination, and then performs a trajectory rationality check on the obtained multiple optimal planned trajectories, and after the check is successful, all the optimal planned trajectories are combined into a corresponding planned trajectory set and saved. Through the present invention, the optimal planned trajectory in any possible scenario in the future time period can be saved as the trajectory data to be checked, which not only avoids the problem of only precipitating one planned trajectory in the conventional processing method, but also provides a complete query basis for real-time trajectory query, thereby improving vehicle safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of a processing method for trajectory planning based on multi-target predicted trajectories provided in Embodiment 1 of the present invention;

[0034] Figure 2 A schematic diagram of a planned trajectory provided in Embodiment 1 of the present invention;

[0035] Figure 3 A schematic diagram of the structure of an electronic device provided in Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] Embodiment 1 of the present invention provides a processing method for trajectory planning based on multi-target predicted trajectories, such as Figure 1 As shown in the schematic diagram of a processing method for trajectory planning based on multi-target predicted trajectories provided in the first embodiment of the present invention, the method mainly includes the following steps:

[0038] Step 1: The planning module receives the multi-target prediction trajectory set output by the upstream prediction module at any time t0;

[0039] Among them, the predicted trajectory period of the multi-target predicted trajectory set is the future period T: [t0, t0+L], t0 is the start time of the period, t0+L is the end time of the period, and L is the duration of the period; the multi-target predicted trajectory set includes multiple first target trajectory groups R; the first target trajectory group R includes multiple first target trajectory p i,j , i is the target index, j is the trajectory index of target i, 1≤i≤n, 1≤j≤m i , n is the number of targets, m i is the number of trajectories of target i; each first target trajectory p i,j is a possible motion trajectory of the corresponding target i in the future period T; any two first target trajectories p in each first target trajectory group R i,j There is an overlapping trajectory between them starting from the starting position.

[0040] Here, the multi-target predicted trajectory set is a complete set of multiple possible motion trajectories of all obstacles (targets) in the traffic environment where the vehicle is located in the future period T; each first target trajectory group R corresponds to a target i, and each first target trajectory p in the trajectory group i,j A possible motion trajectory corresponding to the target i; in the embodiment of the present invention, the time points at which the multiple possible trajectory switching (mutation) of each target in the future period T occurs are all after the starting time point t0, so the multiple first target trajectories p corresponding to the same target i,j There must be a section of the trajectory of the head position that overlaps, so any two first target trajectories p in each first target trajectory group R i,j There is an overlapping trajectory starting from the starting position (the trajectory position corresponding to time t0).

[0041] For example, there are two obstacle targets around the vehicle, namely vehicles 1 and 2. The upstream prediction module predicts the time t in the future period T. a Vehicle 1 has two possible movements (going straight, turning left), while vehicle 2 has only one possible movement (going straight) in the entire process of the future period T; then, the number of targets n = 2, the number of trajectories m for target i = 1 i=1 = 2, the number of trajectories of target i = 2 is m i=2 =1, that is, the multi-target prediction trajectory set should include two first target trajectory groups R1 and R2, and R1 has two first target trajectories p 1,1 、p 1,2 , R2 has 1 first target trajectory p 2,1 , the first target trajectory p in the first target trajectory group R1 1,1 、p 1,2 In the period [t0,t a ]’s trajectories overlap.

[0042] Step 2: Based on the multi-target predicted trajectory set, any possible predicted trajectory of all targets in the future period T is combined and each combination is regarded as a possible future scenario, thereby obtaining a scenario trajectory combination f corresponding to multiple future scenarios. x ;

[0043] Specifically, it includes: selecting a first target trajectory p from each first target trajectory group R of the multi-target trajectory set i,j Combine them into the corresponding scene trajectory combination f x ;

[0044] Where x is the scene index, 1≤x≤w, w is the number of scenes; the scene trajectory combination f x The first target trajectory p of the number of targets n i,j Composition, and each first target trajectory p in the combination i,j The target index i is not repeated; each scene trajectory combination f x Each corresponds to a possible future scene x in the future period T; any combination of two scene trajectories f x The first target trajectory p in each combination i,j The combination relationship is not repeated; the number of scenes is a combination function,

[0045] For example, it is known that the multi-target predicted trajectory set in the future period T includes two first target trajectory groups R1 and R2, and R1 has two first target trajectories p 1,1 、p 1,2 , R2 has 1 first target trajectory p 2,1 ;

[0046] Then, select a first target trajectory p from each first target trajectory group R of the multi-target trajectory set i,j By combining them, the number of scenes can be calculated The corresponding two scene trajectory combinations f x=1 、f x=2 is: f1(p 1,1 , p 1,2 )、f2(p 1,2 , p 1,2 ).

[0047] Step 3: Combine f according to the trajectories of each scene x Plan the optimal driving trajectory of the vehicle in the corresponding scene x in the future period T to obtain the corresponding optimal planning trajectory d x ;

[0048] Specifically, it includes: step 31, obtaining the starting reference position and the target reference position of the vehicle in the future time period T;

[0049] Here, the starting reference position and target reference position of the vehicle in the future time period T are actually the current position coordinates of the vehicle and the position coordinates of the destination of the vehicle. These two position information can be obtained from the positioning module and map module of the vehicle's automatic driving system.

[0050] Step 32, when the combination of the trajectory and the current scene is not satisfied x Any first target trajectory p i,j Under the condition of a collision, the vehicle's driving trajectory from the starting reference position to the target reference position in the future period T is planned to obtain one or more planned trajectories d x,a ;

[0051] Where a is the planning trajectory index of scene x, 1≤a;

[0052] Here, the planning algorithm for driving trajectory planning is similar to the path planning and motion state planning algorithms used in the conventional planning module, and will not be further described here; it should be noted that the current step 32 is to plan the corresponding vehicle trajectory for each multi-target motion scene that may appear in the future, and each scene trajectory combination f x Corresponding to a possible future scenario and including a possible motion trajectory of all obstacle targets (such as vehicles 1 and 2) in the corresponding scenario; and multiple planned trajectories may appear during trajectory planning. x,a of;

[0053] Step 33, plan the trajectory d x,a Counting the number of , generates a corresponding first number;

[0054] For example, Figure 2 As shown in the schematic diagram of the planning trajectory provided in the first embodiment of the present invention, it is known that there are two obstacle targets around the vehicle, namely, vehicles 1 and 2. In the future time period T, t a Vehicle 1 has two possible movements (going straight and turning left), while vehicle 2 has only one possible movement (going straight) in the whole process of future period T. Then the multi-target prediction trajectory set in the future period T includes two first target trajectory groups R1 and R2. R1 has two first target trajectories p 1,1 、p 1,2 , R2 has 1 first target trajectory p 2,1 , we get the two scene trajectory combinations f x=1 、f x=2 is: f1(p 1,1 , p 1,2 )、f2(p 1,2 , p 1,2 ); Let f1(p 1,1 , p 1,2 ) When planning, two planning trajectories are obtained, namely d 1,1 d 1,2 , based on f2(p 1,2 , p 1,2 ) When planning, a planning trajectory is obtained as d 2,1 ;

[0055] Then, in the current scene trajectory combination f x is f1(p 1,1 , p 1,2 ), the corresponding planning trajectory d x,a The number of is the first number is 2; in the current scene trajectory combination f x is f2(p 1,2 , p 1,2 ), the corresponding planning trajectory d x,a The number of, that is, the first number is 1;

[0056] Step 34, identify the first number; if the first number is 1, the only planned trajectory d x,a As the optimal planning trajectory d corresponding to the current scene x x Output; if the first number is greater than 1, then the preset trajectory cost calculation principle is used to calculate each planned trajectory d x,a The trajectory cost is calculated to generate the corresponding first cost data, and the planning trajectory d corresponding to the first cost data with the smallest value is x,a As the optimal planning trajectory d corresponding to the current scene x x Output.

[0057] Here, the embodiment of the present invention outputs only one optimal planning trajectory in each scenario x. The optimal trajectory screening principle here is the preset trajectory cost calculation principle, which has multiple setting standards and implementation methods; one of which can be the principle of taking the overall displacement distance of the trajectory as the minimum value, that is, based on each planning trajectory d x,a The overall displacement distance is used to calculate the corresponding trajectory cost, and the larger the displacement, the greater the cost. Finally, the planning trajectory d corresponding to the first cost data with the smallest value is taken. x,a That is, the planned trajectory d is the minimum displacement distance of the trajectory x,a As the optimal planning trajectory d x Output; Another one can be the principle of minimizing the overall energy consumption of the trajectory, that is, based on each planned trajectory d x,a The energy consumption of the overall trajectory is used to calculate the corresponding trajectory cost, and the greater the energy consumption, the greater the cost. Finally, the planning trajectory d corresponding to the first cost data with the minimum value is taken x,a That is, the planned trajectory d that minimizes the overall trajectory energy consumption x,a As the optimal planning trajectory d x Output.

[0058] by Figure 2 For example, in the current scene trajectory combination f x is f1(p 1,1 , p 1,2 ) and obtain two corresponding planning trajectories d 1,1 ,d 1,2 , assuming that the overall energy consumption of the trajectory is minimized as the trajectory cost calculation principle for d 1,1 ,d 1,2 The trajectory cost is calculated to get d 1,1 The corresponding first cost data is less than d 1,2 The corresponding first cost data, so choose d 1,1 As f1(p 1,1 , p 1,2 ) of the optimal planning trajectory d1; in the current scene trajectory combination f x is f2(p 1,2 , p 1,2 ) and obtain a planned trajectory d 2,1 , then directly use d 2,1 As f2(p 1,2 , p 1,2 )’s optimal planning trajectory d2.

[0059] Step 4: For all optimal planning trajectories d x Conduct trajectory plausibility checks;

[0060] Specifically, step 41 includes: for all optimal planning trajectories d xConduct trajectory motion rationality check;

[0061] Specifically include: in each optimal planning trajectory d x The previous trajectory point is recorded as the previous trajectory point, and the later trajectory point is recorded as the next trajectory point. The relative displacement of the next trajectory point relative to the previous trajectory point is calculated. If the current optimal planning trajectory d x If all relative displacements of are greater than 0, the corresponding inspection state is set to the passed state; if all optimal planning trajectories d x If all corresponding inspection states are passed, the trajectory motion rationality check is successful;

[0062] Here, the optimal planning trajectory d x It can be understood as a number of trajectory points arranged in chronological order; the current optimal planning trajectory d x All relative displacements are greater than 0, indicating that the current optimal planning trajectory d x From the first trajectory point to the last trajectory point, the position of the vehicle is always changing, that is, the vehicle is always in motion rather than stopped, that is, the current optimal planning trajectory d x It is a normal planning trajectory that can continuously approach the target reference position from the starting position of the vehicle, so the corresponding inspection state is set to the passing state at this time; all optimal planning trajectories d x The corresponding inspection status is all passed, indicating that all optimal planning trajectories d x They are all normal planned trajectories, and the trajectory motion rationality check is successful;

[0063] Step 42: If the trajectory motion direction check is successful, then for any two optimal planning trajectories d x Conduct reasonableness check on trajectory branching time;

[0064] Specifically include: Step 421, select two optimal planning trajectories d x The scene trajectories corresponding to the first and second trajectories are combined into f x They are recorded as the corresponding first and second combinations respectively;

[0065] by Figure 2 For example, if there are only two optimal planning trajectories d1 and d2, the first and second trajectories are d1 and d2; the corresponding first and second combinations are f1(p 1,1 , p 1,2 )、f2(p 1,2 , p 1,2 );

[0066] Step 422: In the first and second combinations, the two first target trajectories p corresponding to each target i are i,jConfirm the overlapped trajectory starting from the starting position and use the confirmed overlapped trajectory end time as the corresponding first target trajectory branch time t 1,i , so that n first target trajectory branch times t can be obtained 1,i ; and select the earliest first target trajectory branch time t 1,i As the first branch time t1 corresponding to the first and second combinations;

[0067] Here, it should be noted that if the two first target trajectories p of target i in the first and second combinations i,j If the same, the corresponding first target trajectory branch time t 1,i It should be the end time of the trajectory (t0+L);

[0068] by Figure 2 For example, there are two targets, vehicles 1 and 2, and the first and second combinations are f1(p 1,1 , p 1,2 )、f2(p 1,2 , p 1,2 ); The two first target trajectories corresponding to vehicle 1 in the first and second combinations are p 1,1 、p 1,2 , p of vehicle 1 1,1 、p 1,2 The previous trajectory overlaps and starts from the trajectory point time t a Then the first target trajectory branch time of vehicle 1 is t 1,i=1 =t a ; The two first target trajectories corresponding to vehicle 2 in the first and second combinations are both p 2,1 , then the first target trajectory branch time t of vehicle 2 1,i=2 It should be the trajectory end time (t0+L); because the final two first target trajectory branch times t 1,i=1 ,t 1,i=2 Medium 1,i=1 Earlier, so take t 1,i=1 =t a As the first branch time t1;

[0069] Step 423, confirming the overlapping trajectory of the first and second trajectories starting from the starting position and using the confirmed ending time of the overlapping trajectory as the corresponding second branching time t2;

[0070] by Figure 2 For example, if there are only two optimal planning trajectories d1 and d2, the first and second trajectories are d1 and d2. It is known that the previous trajectories of d1 and d2 overlap and the time from the trajectory point is t b The second branch time t2 = t b ;

[0071] Step 424, calculate the time difference Δt = t2 - t1 between the second branch time t2 and the first branch time t1; and identify whether the time difference Δt satisfies a preset reasonable time difference range. If the time difference Δt satisfies the reasonable time difference range, the trajectory branch time rationality check is successful;

[0072] Among them, the reasonable time difference range is a time difference range greater than 0 and less than or equal to a specified threshold δ;

[0073] Here, in the embodiments of the present invention, t1 < t2 ≤ (t1 + δ) is actually used to constrain the planned trajectory branch time and the predicted trajectory branch time, that is, it is required that the branch time of the planned trajectory should be later than the predicted trajectory branch time; the specified threshold δ is a system delay parameter, which is default set according to the overall perception delay of the vehicle's own perception sensors and perception modules;

[0074] Step 43, if the trajectory branch time check is successful, the trajectory rationality check is successful.

[0075] Step 5, if the trajectory rationality check is successful, then save the set of planned trajectories composed of all optimal planned trajectories d x formed.

[0076] Figure 3 It is a schematic structural diagram of an electronic device provided in Embodiment 2 of the present invention. The electronic device can be the aforementioned terminal device or server, or a terminal device or server connected to the aforementioned terminal device or server to implement the method of the embodiments of the present invention. As Figure 3 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 actions of the transceiver 303. Various instructions may be stored in the memory 302 to complete various processing functions and implement the processing steps described in the foregoing method embodiments. Preferably, the electronic device involved in the embodiments 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 aforementioned communication port 306 is used for the electronic device to connect and communicate with other peripherals.

[0077] In Figure 3 The system bus 305 mentioned can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The system bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3The term "communication interface" is represented by only one thick line, but it does not mean that there is only one bus or one type of bus. The communication interface is used to realize the communication between the database access device and other devices (such as clients, read-write libraries, and read-only libraries). The memory may include random access memory (RAM) and may also include non-volatile memory (Non-Volatile Memory), such as at least one disk storage.

[0078] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a graphics processing unit (GPU), etc.; it can 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.

[0079] It should be noted that an embodiment of the present invention further provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the method and processing process provided in the above embodiments.

[0080] An embodiment of the present invention further provides a chip for executing instructions, wherein the chip is used to execute the processing steps described in the aforementioned method embodiment.

[0081] The embodiment of the present invention provides a processing method, electronic device and computer-readable storage medium for trajectory planning based on multi-target predicted trajectories. The planning module first confirms the scene trajectory combination of all possible multi-target motion scenes in the future time period based on the multi-target predicted trajectory set sent by the upstream prediction module, and then plans the optimal driving trajectory of the vehicle in each scene in the future time period according to each scene trajectory combination, and then performs a trajectory rationality check on the obtained multiple optimal planned trajectories, and after the check is successful, all the optimal planned trajectories are combined into a corresponding planned trajectory set and saved. Through the present invention, the optimal planned trajectory in any possible scenario in the future time period can be saved as the trajectory data to be checked, which not only avoids the problem of only precipitating one planned trajectory in the conventional processing method, but also provides a complete query basis for real-time trajectory query, thereby improving vehicle safety.

[0082] The professionals should further realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to the function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0083] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0084] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A processing method for trajectory planning based on multi-target predicted trajectories, characterized in that: The method comprises: The planning module receives the multi-target prediction trajectory set output by the upstream prediction module at any time t0; the prediction trajectory period of the multi-target prediction trajectory set is the future period T: [t0, t0+L], t0 is the start time of the period, t0+L is the end time of the period, and L is the duration of the period; the multi-target prediction trajectory set includes multiple first target trajectory groups R; the first target trajectory group R includes multiple first target trajectory p i,j , i is the target index, j is the trajectory index of target i, 1≤i≤n, 1≤j≤m i , n is the number of targets, m i is the number of trajectories of target i; According to the multi-target predicted trajectory set, any possible predicted trajectory of all targets in the future period T is combined and each combination is regarded as a possible future scene, so as to obtain scene trajectory combinations f corresponding to multiple future scenes respectively. x ; x is the scene index, 1≤x≤w, w is the number of scenes; the scene trajectory combination f x The first target trajectory p of the target number n i,j The first target trajectory p in the combination i,j The target index i is not repeated; According to each of the scene trajectory combinations f x Plan the optimal driving trajectory of the vehicle in the corresponding scene x in the future period T to obtain the corresponding optimal planning trajectory d x ; For all the optimal planning trajectories d x Conduct trajectory plausibility checks; If the trajectory rationality check succeeds, then all the optimal planned trajectories d x The planned trajectory set composed of the above two components is saved.

2. The processing method for trajectory planning based on multi-target predicted trajectories according to claim 1, characterized in that: Each of the first target trajectories p i,j is a possible motion trajectory of the corresponding target i in the future period T; any two of the first target trajectories p in each of the first target trajectory groups R i,j There is an overlapping trajectory between them starting from the starting position.

3. The processing method for trajectory planning based on multi-target predicted trajectories according to claim 1 is characterized in that: According to the multi-target predicted trajectory set, any possible predicted trajectory of all targets in the future period T is combined and each combination is regarded as a possible future scene, so as to obtain scene trajectory combinations f corresponding to multiple future scenes respectively. x , specifically including: Select one of the first target trajectories p from each of the first target trajectory groups R of the multi-target trajectory set i,j Combine to form the corresponding scene trajectory combination f x ; Among them, each of the scene trajectory combinations f x Each of them corresponds to a possible future scene x in the future period T; any two scene trajectory combinations f x The first target trajectory p in each combination i,j The combination relationship is not repeated; the number of scenes is a combination function.

4. The processing method for trajectory planning based on multi-target predicted trajectories according to claim 1 is characterized in that: The combination of each scene trajectory f x Plan the optimal driving trajectory of the vehicle in the corresponding scene x in the future period T to obtain the corresponding optimal planning trajectory d x , specifically including: Obtaining the starting reference position and the target reference position of the vehicle in the future time period T; In the case where the trajectory combination f is not satisfied x Any of the first target trajectories p i,j Under the condition of a collision, the driving trajectory of the vehicle from the starting reference position to the target reference position in the future period T is planned to obtain one or more planned trajectories d x,a ; a is the planning trajectory index of scene x, 1≤a; For the planned trajectory d x,a Counting the number of , generates a corresponding first number; Identify the first number; if the first number is 1, then the unique planned trajectory d x,a As the optimal planning trajectory d corresponding to the current scene x x Output; if the first number is greater than 1, then calculate the cost of each planned trajectory d according to the preset trajectory cost calculation principle x,a The trajectory cost is calculated to generate the corresponding first cost data, and the planned trajectory d corresponding to the first cost data with the smallest value is x,a As the optimal planning trajectory d corresponding to the current scene x x Output.

5. The processing method for trajectory planning based on multi-target predicted trajectories according to claim 1, characterized in that: For all the optimal planning trajectories d x Perform trajectory rationality checks, including: For all the optimal planning trajectories d x Perform trajectory motion rationality check; if the trajectory motion direction check is successful, then for any two of the optimal planning trajectories d x Perform a trajectory branch time rationality check; if the trajectory branch time check is successful, then the trajectory rationality check is successful.

6. The processing method for trajectory planning based on multi-target predicted trajectories according to claim 5, characterized in that: For all the optimal planning trajectories d x Perform trajectory motion rationality check, including: In each of the optimal planning trajectories d x The previous trajectory point is recorded as the previous trajectory point, and the next trajectory point is recorded as the next trajectory point. The relative displacement of the next trajectory point relative to the previous trajectory point is calculated. If the current optimal planning trajectory d x If all the relative displacements are greater than 0, the corresponding inspection state is set to the pass state; if all the optimal planning trajectories d x If all the corresponding inspection states are passed, the trajectory motion rationality check is successful.

7. The processing method for trajectory planning based on multi-target predicted trajectories according to claim 5, characterized in that: For any two of the optimal planning trajectories d x Perform a reasonableness check on the trajectory branch time, including: Choose any two of the optimal planning trajectories d x The scene trajectories corresponding to the first and second trajectories are combined into f x They are recorded as the corresponding first and second combinations respectively; In the first and second combinations, the two first target trajectories p corresponding to each target i i,j Confirm the overlapped trajectory starting from the starting position and use the confirmed overlapped trajectory end time as the corresponding first target trajectory branch time t 1,i , so that n branch times t of the first target trajectory can be obtained 1,i ; and select the earliest first target trajectory branch time t 1,i as a first branch time t1 corresponding to the first and second combinations; Confirming the overlapping trajectory of the first and second trajectories starting from the starting position and using the confirmed ending time of the overlapping trajectory as the corresponding second branching time t2; Calculate the time difference △t=t2-t1 between the second branch time t2 and the first branch time t1; and identify whether the time difference △t satisfies the preset reasonable time difference range. If the time difference △t satisfies the reasonable time difference range, the trajectory branch time rationality check is successful; the reasonable time difference range is a time difference range greater than 0 and less than or equal to the specified threshold δ.

8. An electronic device, characterized in that: include: memory, processors, and transceivers; The processor is used to couple with the memory, read and execute instructions in the memory, so as to implement the method steps described in any one of claims 1 to 7; The transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, which, when executed by a computer, enable the computer to execute the method according to any one of claims 1 to 7.