Track planning method, track display method, device, vehicle, medium and chip

Abstract

The embodiment of the invention provides a trajectory planning method, a trajectory display method, equipment, a vehicle, a medium and a chip, and relates to the field of automatic driving. The trajectory planning method comprises the following steps: acquiring map element information in a target area and current position information and historical position information of a target object according to environmental perception data of the target area; inputting the current position information, the historical position information and the map element information into a feature extraction fusion model to obtain a first fusion feature output by the feature extraction fusion model; and generating a target planning trajectory of the target vehicle according to the first fusion feature. Wherein the target area is an area containing the position of a target vehicle with a to-be-planned trajectory, the target object comprises the target vehicle and an obstacle, and the first fusion feature comprises interaction features representing mutual influence between different target objects and between the target object and map elements.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and more specifically, to a trajectory planning method, a trajectory display method, a device, a vehicle, a medium, and a chip. Background Art

[0002] With the increasing application of autonomous driving technology, the environment faced by autonomous vehicles during driving is becoming more and more complex. In order to safely and reliably control the operation of autonomous vehicles, it is necessary to plan the trajectory of the vehicle. However, the trajectory planning methods of related technologies have poor robustness in complex environments. How to improve the robustness of vehicle trajectory planning has become an urgent problem to be solved. Summary of the invention

[0003] In view of this, the embodiments of the present disclosure propose a new technical solution for trajectory planning.

[0004] According to a first aspect of an embodiment of the present disclosure, a trajectory planning method is provided, the method comprising:

[0005] According to the environmental perception data of the target area, map element information in the target area and current position information and historical position information of the target object are obtained; wherein the target area is an area including the position of the target vehicle whose trajectory is to be planned, and the target object includes the target vehicle and obstacles;

[0006] Inputting the current location information, the historical location information and the map element information into a feature extraction fusion model to obtain a first fusion feature output by the feature extraction fusion model; wherein the first fusion feature includes an interaction feature characterizing the mutual influence between different target objects and between target objects and map elements;

[0007] A target planning trajectory of the target vehicle is generated according to the first fusion feature.

[0008] Optionally, the feature extraction fusion model includes a target encoder and a target decoder; the feature extraction fusion model is used to generate the first fusion feature in the following manner:

[0009] Based on the target encoder, feature extraction is performed on the map element information, the current position information and the historical position information to obtain map information features, historical trajectory features and first predicted trajectory features of the target object;

[0010] Based on the target decoder, feature fusion is performed on the map information feature, the historical trajectory feature of the target object and the first predicted trajectory feature to obtain the first fused feature.

[0011] Optionally, the target decoder is a decoder based on a cross attention mechanism; the target decoder performs feature fusion on the map information feature, the predicted trajectory feature and the historical trajectory feature of the target object to obtain the first fused feature, including:

[0012] Based on the cross attention mechanism, the first predicted trajectory feature and the historical trajectory feature are integrated to obtain a second predicted trajectory feature;

[0013] The second predicted trajectory feature and the map information feature are integrated based on a cross attention mechanism to obtain a third predicted trajectory feature;

[0014] The third predicted trajectory feature and the historical trajectory feature are fused based on a cross attention mechanism to obtain the first fused feature.

[0015] Optionally, the target decoder includes a first cross attention unit; the fusing the first predicted trajectory feature and the historical trajectory feature based on the cross attention mechanism to obtain a second predicted trajectory feature includes:

[0016] Generate a first query feature according to the first predicted trajectory feature and the target object relative time dimension interaction feature; the target object relative time dimension interaction feature is an interaction feature for characterizing the target object in the time dimension obtained by embedding and fusing the current position information and the historical position information of the target object;

[0017] Generate a first key feature and a first value feature according to the historical trajectory feature and the target object relative time dimension interaction feature;

[0018] The first query feature, the first key feature, and the first value feature are input into a first cross attention unit to generate the second predicted trajectory feature.

[0019] Optionally, the target decoder further includes a first cross attention unit; the fusing the second predicted trajectory feature and the map information feature based on the cross attention mechanism to obtain a third predicted trajectory feature includes:

[0020] generating a second query feature according to the second predicted trajectory feature and the target object relative position interaction feature; the target object relative position interaction feature is a feature obtained by embedding and fusing the current position information and the historical position information of the target object and used to characterize the mutual influence between the target objects;

[0021] generating a second key feature and a second value feature according to the map information feature and the target object relative position interaction feature;

[0022] The second query feature, the second key feature and the second value feature are input into a second cross attention unit to generate the third predicted trajectory feature.

[0023] Optionally, the fusing the third predicted trajectory feature and the historical trajectory feature based on the cross attention mechanism to obtain the first fused feature includes:

[0024] generating a third query feature according to the third predicted trajectory feature and the target object relative map interaction feature; the target object relative map interaction feature is a feature obtained by embedding and fusing the current position information, the historical position information and the map element information of the target object and used to characterize the mutual influence between the target object and the map element;

[0025] generating a third key feature and a third value feature according to the historical trajectory feature and the target object relative map interaction feature;

[0026] The third query feature, the third key feature and the third value feature are input into a third cross attention unit to generate the first fusion feature.

[0027] Optionally, generating a target planning trajectory of the target vehicle according to the first fusion feature includes:

[0028] Inputting the first fusion feature into a pre-generated trajectory prediction planning model to obtain an initial planning trajectory of the target vehicle and an obstacle prediction trajectory of the obstacle output by the trajectory prediction planning model;

[0029] A target planned trajectory of the target vehicle is determined according to the initial planned trajectory and the obstacle predicted trajectory.

[0030] Optionally, the trajectory prediction planning model includes a self-vehicle trajectory planning module and an obstacle trajectory prediction module; the trajectory prediction planning model generates the initial planning trajectory of the target vehicle and the obstacle prediction trajectory of the obstacle in the following manner:

[0031] The ego vehicle trajectory planning module generates an initial planning trajectory of the target vehicle according to the first fusion feature;

[0032] The obstacle trajectory prediction module generates an obstacle prediction trajectory of the obstacle according to the first fusion feature, the initial planning trajectory and the current position information of the obstacle.

[0033] Optionally, the ego vehicle trajectory planning module includes a first self-attention unit and a first neural network unit; the ego vehicle trajectory planning module generates an initial planning trajectory of the target vehicle according to the first fusion feature, including:

[0034] Extracting a first trajectory feature of the target vehicle from the first fused feature based on a first self-attention unit;

[0035] The first trajectory feature and the current state information of the target vehicle are input into a first neural network unit to obtain an initial planned trajectory of the target vehicle.

[0036] Optionally, the obstacle trajectory prediction module includes a fourth cross-attention unit, a second self-attention unit and a second neural network unit; the obstacle trajectory prediction module generates an obstacle prediction trajectory of the obstacle according to the first fusion feature, the initial planning trajectory and the current position information of the obstacle, including:

[0037] fusing the first fused feature, the initial planned trajectory, and the current position information of the obstacle based on a fourth cross attention unit to obtain a second fused feature;

[0038] Extracting the trajectory feature of the obstacle from the second fused feature based on the second self-attention unit;

[0039] The trajectory features are input into a second neural network unit to obtain an obstacle prediction trajectory of the obstacle.

[0040] Optionally, the fusing the first fused feature, the initial planned trajectory and the current position information of the obstacle based on the fourth cross attention unit to obtain a second fused feature includes:

[0041] After embedding and fusing the current position information of the obstacle and the initial planned trajectory of the target vehicle, the interaction feature of the obstacle relative to the target vehicle is obtained;

[0042] generating a fourth query feature according to the first fusion feature and the obstacle-to-target-vehicle interaction feature;

[0043] Generate a fourth key feature and a fourth value feature according to the initial planned trajectory and the obstacle-to-target vehicle interaction feature;

[0044] The fourth query feature, the fourth key feature and the fourth value feature are input into a fourth cross attention unit to generate the second fusion feature.

[0045] Optionally, determining a target planned trajectory of the target vehicle according to the initial planned trajectory and the obstacle predicted trajectory includes:

[0046] Predicting a collision probability between the target vehicle and an obstacle based on the initial planned trajectory and the obstacle predicted trajectory;

[0047] When the collision probability is greater than or equal to a preset collision probability threshold, updating the initial planned trajectory according to the obstacle prediction trajectory, and using the updated initial planned trajectory as the target planned trajectory to reduce the collision probability;

[0048] When the collision probability is less than a preset collision probability threshold, the initial planning trajectory is used as the target planning trajectory.

[0049] According to a second aspect of an embodiment of the present disclosure, a trajectory display method is provided, the method comprising:

[0050] Determine the target planning trajectory of the target vehicle based on the environmental perception data of the target area;

[0051] Displaying the target planning trajectory;

[0052] Among them, the target planning trajectory is a trajectory generated based on the first fusion feature output by the feature extraction fusion model, the environmental perception data is used to obtain the map element information in the target area and the current position information and historical position information of the target object, the target area is the area including the location of the target vehicle, the target object includes the target vehicle and obstacles, and the feature extraction fusion model is used to output the first fusion feature based on the input current position information, the historical position information and the map element information.

[0053] According to a third aspect of an embodiment of the present disclosure, an electronic device is provided, comprising a memory and a processor, wherein the memory is used to store computer instructions, and the processor is used to call the computer instructions from the memory to execute a method as described in any one of the first aspect and / or the second aspect.

[0054] According to a fourth aspect of an embodiment of the present disclosure, a vehicle is provided, comprising a memory and a processor, wherein the memory is used to store computer instructions, and the processor is used to call the computer instructions from the memory to execute a method as described in any one of the first aspect and / or the second aspect.

[0055] According to a fifth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method described in any one of the first aspect and / or the second aspect is implemented.

[0056] According to a sixth aspect of an embodiment of the present disclosure, a chip is provided, comprising a processing unit, wherein the processing unit is configured to execute a method as described in any one of the first aspect and / or the second aspect.

[0057] Based on the trajectory planning method provided by the embodiment of the present disclosure, a first fusion feature is generated to characterize the mutual influence between different target objects and between the target object and the map element. The target planning trajectory of the target vehicle is generated according to the first fusion feature, and interactive feedback between the target vehicle, obstacles and map elements is realized. Efficient and robust trajectory planning can be achieved in a dynamic environment, thereby improving the safety, stability and driving efficiency of the autonomous driving vehicle.

[0058] Other features and advantages of the present disclosure will become apparent from the following detailed description of exemplary embodiments of the present disclosure with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0060] Figure 1 The present invention is a schematic diagram of an intelligent network connection system to which the method provided by the embodiment of the present disclosure can be applied.

[0061] Figure 2 is based on Figure 1 The illustrated embodiment provides a schematic diagram of a vehicle.

[0062] Figure 3 It is a flowchart of a trajectory planning method provided by an embodiment of the present disclosure.

[0063] Figure 5 It is a structural diagram of a feature extraction fusion model and a trajectory prediction planning model provided by an embodiment of the present disclosure.

[0064] Figure 4 It is a structural diagram of another feature extraction fusion model and trajectory prediction planning model provided in an embodiment of the present disclosure.

[0065] Figure 6 It is a schematic diagram of an encoding and fusion process provided by an embodiment of the present disclosure.

[0066] Figure 7 It is a flowchart of a trajectory display method provided by an embodiment of the present disclosure.

[0067] Figure 8 It is a schematic diagram showing a target planning trajectory provided by an embodiment of the present disclosure.

[0068] Fig. 9 It is a structural schematic diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0069] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure unless otherwise specifically stated.

[0070] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0071] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the above-mentioned technologies, methods, and equipment should be considered as part of the specification.

[0072] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0073] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0074] First, the application scenarios of the embodiments of the present disclosure are described.

[0075] Figure 1 1 is a schematic diagram of an intelligent network system 100 to which the method provided by the embodiment of the present disclosure can be applied. Figure 1 As shown, the intelligent network connection system 100 may include: a vehicle 101, a server 102, and a user terminal 103.

[0076] In some examples, vehicle 101 may be a vehicle with an autonomous driving function. Autonomous driving is also known as unmanned driving or intelligent driving. Vehicles with autonomous driving functions can achieve driving tasks such as environmental perception, decision planning, and control execution. The level of autonomous driving can refer to the automotive intelligence classification standards formulated by the Society of Automotive Engineers (SAE). For example, L0 is manual driving, L1 is assisted driving, L2 is partial autonomous driving, L3 is conditional autonomous driving, L4 is highly autonomous driving, and L5 is fully autonomous driving. The above classification method for autonomous driving levels is only used as an example, and the disclosed embodiments do not limit the classification standards and levels of autonomous driving.

[0077] In some examples, the server 102 may be a single server or a distributed server cluster consisting of multiple servers, and its deployment method may include a local server or a cloud server. The server 102 may communicate with the vehicle 101 and / or the user terminal 103 based on a communication network, and provide various services for the vehicle 101 and / or the user terminal 103. For example, the server may receive the perception data sent by the vehicle, and provide the vehicle with services such as high-precision maps, data analysis, and decision-making planning. For another example, the server may receive query instructions or control instructions sent by the user terminal, and provide corresponding services to the user.

[0078] In some examples, the user terminal 103 may be any form of electronic device that provides services to users, such as a personal computer, a laptop, a smart tablet, a smart phone, a smart wearable device, etc. The user may interact with the vehicle or server through the human-computer interaction terminal configured in the vehicle 101, or may interact with the vehicle or server through the user terminal 103, for example, querying the status and / or parameters of the vehicle through the user terminal, or controlling the vehicle to perform set tasks and / or modify configuration parameters, etc.; wherein the user terminal runs an application based on the intelligent network connection system to achieve interaction with the vehicle or server, and the application may be a local application, a web application, a small program, etc., which is not limited here.

[0079] In some examples, the above-mentioned application running on the user terminal can provide authentication or authorization services for the user. The user who is successfully authenticated and granted the corresponding authority can query and / or control the vehicle within the scope of the granted authority.

[0080] The vehicle 101, the server 102 and the user terminal 103 can communicate with each other through the communication link provided by the communication network 104. The communication network 104 may include one or more networks of any type, for example, the communication network 104 may include the Internet, a local area network (LAN), a wide area network (WAN), a virtual private network (VPN), a public switched telephone network (PSTN), a satellite communication network, Wi-Fi, 2G, 3G, 4G, 5G, 6G, NB-IoT, eMTC, infrared, Bluetooth, NFC and other networks providing communication, or a combination of the above multiple networks. The communication networks between the vehicle 101 and the server 102, between the user terminal 103 and the server 102, and between the user terminal 103 and the vehicle 101 may be the same or different.

[0081] It should be noted that Figure 1The structure of the intelligent network connection system 100 shown in the figure is only illustrative. The intelligent network connection system in the embodiment of the present disclosure is not limited to the above structure, and may include more or fewer devices as needed, and may also combine or split the devices. For example, the intelligent network connection system may not include a user terminal and / or a server; for another example, the user terminal and the server may be deployed together.

[0082] Figure 2 is based on Figure 1 The illustrated embodiment provides a schematic diagram of a vehicle 101. Figure 2 As shown, the vehicle 101 may include a perception component 1011, a computing platform 1012, an execution component 1013, etc. The perception component 1011, the computing platform 1012, and the execution component 1013 may be connected via a bus or other means.

[0083] In some examples, the perception component 1011 can be used to collect information about the vehicle itself or outside. The perception component 1011 may include at least one of a visual sensor unit, a radar, a positioning and navigation unit, an inertial measurement unit (IMU) or other sensing units, wherein the visual sensor unit may include one or more cameras, the radar may include at least one of a laser radar, a millimeter wave radar, an ultrasonic radar or other radars, and the positioning and navigation unit may include at least one of a GPS system, a Beidou system or other global positioning systems.

[0084] In some examples, the computing platform 1012 may include a device with computing capabilities, which is used to process the perception information collected by the perception component 1011 to obtain control information, and send corresponding control instructions to the execution component 1013, so that the execution component 1013 performs corresponding actions, thereby realizing control of the vehicle 101. For example, the computing platform 1012 can perform positioning and mapping (Simultaneous Localization and Mapping, SLAM), path planning, behavior decision-making and other behaviors on the vehicle, thereby realizing autonomous control of the vehicle. The computing platform 1012 may include at least one processor and at least one memory, and each processor may execute instructions stored in the memory individually or collectively to implement the method provided in the embodiment of the present disclosure. The processor in the embodiment of the present disclosure may include a central processing unit (CPU), a graphic processing unit (GPU), a neural-network processing unit (NPU), a tensor processing unit (TPU), a data processing unit (DPU), a digital signal processor (DSP), a field programmable gate array (FPGA), a system on chip (SOC), an application specific integrated circuit (ASIC), a microcontroller unit (MCU) or at least one of other processors. The memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a disk or an optical disk. In addition to storing instructions, the memory can also store data, such as high-precision maps, path information, vehicle location, direction, speed and other data. The data stored in the memory can be retrieved and used by the processor.

[0085] In some examples, the vehicle's computing platform can independently perform computing tasks or communicate with a server to complete computing tasks. For example, the vehicle's computing platform can cooperate with the server to complete corresponding computing tasks.

[0086] The computing platform 1012 may be arranged in the vehicle 101, and part or all of the computing platform 1012 may also be arranged in a server corresponding to the vehicle. For example, some functions of the computing platform 1012 with higher real-time requirements may be arranged in the vehicle, and other functions with lower real-time requirements may be arranged in the server corresponding to the vehicle.

[0087] In some examples, the execution component 1013 is used to execute corresponding actions based on the control of the computing platform 1012, so that the vehicle 101 completes the moving task. The execution component 1013 may include, for example, a power component, a brake component, a transmission component, a steering component, and the like.

[0088] It should be noted that Figure 2 The structure of the vehicle 101 shown in the figure is only illustrative. The vehicle in the embodiment of the present disclosure is not limited to the above structure, and may include more or fewer components as needed, and the devices may be combined or split. For example, the vehicle may not include the above computing platform. For another example, the vehicle may also include a communication component, an interface component, a multimedia component, an input component, an output component, etc.

[0089] The disclosed embodiments can be applied to the self-driving vehicle trajectory planning scenario. With the increasing application of autonomous driving technology, the driving environment faced by autonomous driving vehicles is becoming more and more complex. In order to safely and reliably control the operation of autonomous driving vehicles, on the basis of global path planning, short-term trajectory planning is also required for the vehicle to determine the best driving route for the vehicle itself.

[0090] In the related art, the obstacle trajectory prediction module and the vehicle trajectory planning module can be processed as independent modules respectively. The trajectory prediction module can predict the future motion state of obstacles around the vehicle based on the perception information, while the vehicle trajectory planning module plans the trajectory of the vehicle based on the predicted future motion state of obstacles and the global path planning to determine the best driving route for the vehicle. However, this processing method of separating prediction from planning may lead to inconsistency between prediction results and planning decisions when facing complex dynamic environments, because trajectory prediction and trajectory planning are independent of each other and lack real-time interactive feedback, thereby affecting the overall response speed and decision-making accuracy of the system. Secondly, most of the existing trajectory planning methods are based on static or semi-static environments, which are difficult to effectively respond to the behavioral changes of other dynamic traffic participants, and may cause problems such as inability to accurately avoid obstacles, uneven routes or unstable decisions. In addition, with the increasing complexity of urban road scenes and highway scenes, the traditional separate processing method shows certain limitations in computing resources and time costs, and it is difficult to meet the real-time requirements of autonomous driving systems. Therefore, there is an urgent need for a new trajectory planning method that can achieve efficient and robust trajectory planning in a dynamic environment.

[0091] Figure 3 is a flow chart of a trajectory planning method provided by an embodiment of the present disclosure. The trajectory planning method can be Figure 1 The vehicle and / or server are executed as shown. Figure 3 As shown, the trajectory planning method of this embodiment may include the following steps S310 to S330.

[0092] Step S310, obtaining map element information in the target area and current location information and historical location information of the target object according to the environmental perception data of the target area.

[0093] The target area may be an area containing the location of the target vehicle whose trajectory is to be planned. For example, the target area may be the maximum area detected by the sensor of the target vehicle, or the target area may be an area within a preset range around the location of the target vehicle. The target objects within the target area may include the target vehicle and obstacles, and the obstacles may be one or more. Optionally, the obstacles may be dynamic obstacles, such as other vehicles or pedestrians, or static obstacles, such as signs, cones, etc.

[0094] The map element information may be a representation of static element information in the target area in the high-precision map. For example, the map element information may include road information, lane information, intersection information, traffic sign information, sign line information, arrow information, and other information in the high-precision map.

[0095] The current location information of the target object may be location information at the current moment, for example, may be location information determined based on environmental perception data at the current moment.

[0096] The historical location information of the target object may include multiple location information determined based on the environmental perception data at multiple historical moments, for example, the movement trajectory of the target object within a preset time (such as 3 seconds). For static obstacles, the movement trajectory may be a location point, and for dynamic obstacles, the movement trajectory may include a straight line or curve composed of multiple location points.

[0097] In some examples, the above-mentioned environmental perception data may be data obtained by detecting the surrounding environment of the vehicle through sensors during the driving of the vehicle. For example, the environmental perception data may include image data collected by visual sensors and / or point cloud data collected by radars. The environmental perception data may be input into a pre-generated neural network model to obtain map element information within the target area output by the neural network model and current location information and historical location information of the target object.

[0098] Step S320, inputting the current position information, the historical position information and the map element information into a feature extraction fusion model to obtain a first fusion feature output by the feature extraction fusion model.

[0099] The first fusion feature may include an interaction feature characterizing the mutual influence between different target objects and between the target object and the map element.

[0100] For example, the feature extraction fusion model can extract features from the input current position information, historical position information, and map element information of the target object including the target vehicle and obstacles, and output a first fusion feature that integrates the information related to the target vehicle, obstacles, and map elements, and the interactive features of the mutual influence between different target objects and between the target objects and map elements in the first fusion feature, so as to form an overall understanding of the dynamic environment based on the first fusion feature. Optionally, the first fusion feature can be a high-dimensional fusion feature, such as 128 dimensions or 256 dimensions.

[0101] Step S330: generating a target planning trajectory of the target vehicle according to the first fusion feature.

[0102] For example, the first fusion feature may be input into a pre-generated trajectory prediction planning model to obtain a target planning model.

[0103] By adopting the method of steps S310 to S330 above, a first fusion feature is generated to characterize the mutual influence between different target objects and between the target object and the map elements, and a target planning trajectory of the target vehicle is generated according to the first fusion feature, so as to realize interactive feedback between the target vehicle, obstacles and map elements, and realize efficient and robust trajectory planning in a dynamic environment, thereby improving the safety, stability and driving efficiency of the autonomous driving vehicle.

[0104] In some embodiments of the present disclosure, the above step S330 may include the following steps S331 and S332:

[0105] Step S331, inputting the first fusion feature into a pre-generated trajectory prediction planning model to obtain an initial planning trajectory of the target vehicle and an obstacle prediction trajectory of the obstacle output by the trajectory prediction planning model.

[0106] In some examples, the trajectory prediction planning model may include a network model using a multilayer perceptron (MLP), through which the initial planning trajectory of the target vehicle and the obstacle prediction trajectory of the obstacle can be predicted from the first fused feature. The obstacle prediction trajectory includes the predicted position of the obstacle at a future time, which can provide input basis for further planning and obstacle avoidance decisions of the target vehicle.

[0107] Step S332, determining the target planned trajectory of the target vehicle according to the initial planned trajectory and the obstacle predicted trajectory.

[0108] The target planning trajectory and the initial planning trajectory may be the same or different. For example, the initial planning trajectory may be used as the target planning trajectory, or the target planning trajectory may be obtained by adjusting the initial planning trajectory according to the obstacle prediction trajectory.

[0109] In some examples, the collision probability of the target vehicle colliding with the obstacle can be predicted based on the initial planned trajectory and the obstacle prediction trajectory; when the collision probability is greater than or equal to a preset collision probability threshold, the initial planned trajectory is updated according to the obstacle prediction trajectory, and the updated initial planned trajectory is used as the target planned trajectory to reduce the collision probability; or, when the collision probability is less than the preset collision probability threshold, the initial planned trajectory is used as the target planned trajectory.

[0110] The collision probability may be used to indicate the probability of a collision between the target vehicle and the obstacle at a future time. The preset collision probability threshold may be any pre-set value.

[0111] In other examples, a safety assessment may be performed on the target planned trajectory. The assessment indicators may include collision probability, trajectory curvature, trajectory comfort, etc. If the assessment result meets preset conditions, the target vehicle may be controlled according to the target planned trajectory.

[0112] In this way, interactive feedback between vehicle trajectory planning and obstacle trajectory prediction can be achieved, and efficient and robust trajectory planning can be realized in dynamic environments, thereby improving the safety, stability and driving efficiency of autonomous driving vehicles.

[0113] Figure 4 is a structural diagram of a feature extraction fusion model and a trajectory prediction planning model provided by an embodiment of the present disclosure. , such as Figure 4 As shown, the feature extraction fusion model 41 may include a target encoder 411 and a target decoder 412; the trajectory prediction planning model 42 may include an ego vehicle trajectory planning head 421 and an obstacle trajectory prediction head 422, the ego vehicle trajectory planning head may also be referred to as an ego vehicle trajectory planning module, and the obstacle trajectory prediction head may also be referred to as an obstacle trajectory prediction module.

[0114] In some examples, the feature extraction fusion model can be used to generate a first fused feature in the following manner: based on the target encoder, feature extraction is performed on map element information, current position information, and historical position information to obtain map information features, historical trajectory features of the target object, and first predicted trajectory features; based on the target decoder, feature fusion is performed on the map information features, historical trajectory features of the target object, and the first predicted trajectory features to obtain a first fused feature.

[0115] In some examples, the target encoder 411 may be any encoder generated by pre-training, for example, a PnP Encoder (Prediction and Planning Encoder). Optionally, the target encoder may include a convolution unit and an attention mechanism unit to extract features of the input map element information, current location information, and historical location information to obtain high-dimensional (e.g., 128-dimensional) map information features, historical trajectory features of the target object, and first predicted trajectory features.

[0116] For example, the above-mentioned map element information input to the target encoder can be expressed as (B, M, P, 2), where B represents a data batch, M represents the number of map elements, P represents the number of key points of the map element, and 2 represents that the coordinates of each key point are expressed in the form of two-dimensional coordinates. The current position information of the above-mentioned target object can be expressed as (B, N, 1, 2), where B represents a data batch, N represents the number of target objects, 1 represents the current frame or current moment, and 2 represents that the current position coordinates of the target object are expressed in the form of two-dimensional coordinates. The current position information of the above-mentioned target object can be expressed as (B, N, HT, 2), where B represents a data batch, N represents the number of target objects, HT represents the number of historical moments or historical frames, and 2 represents that the historical position coordinates of the target object are expressed in the form of two-dimensional coordinates. It should be noted that the key point coordinates of the above-mentioned map elements, the current position coordinates and the historical position coordinates of the target object can also be expressed in the form of three-dimensional coordinates.

[0117] The above-mentioned map information features output by the target encoder can be expressed as (B*M*K, 1, M, 128), where B represents a data batch, M represents the number of map elements, K represents the maximum number of pre-set predicted trajectories, such as 6 or 8, and 128 represents that the dimension after encoding is 128 dimensions. The first predicted trajectory features of the above-mentioned target object can be expressed as (B, K, N, 128), where B represents a data batch, K represents the maximum number of pre-set predicted trajectories, N represents the number of target objects, and 128 represents that the dimension after encoding is 128 dimensions. The historical trajectory features of the above-mentioned target object can be expressed as (B, N, HT, 128), where B represents a data batch, N represents the number of target objects, HT represents the number of historical moments or historical frames, and 128 represents that the dimension after encoding is 128 dimensions.

[0118] In this way, the target encoder can obtain high-dimensional feature representations of map information, current location information and historical location information of the target object, and extract richer features to improve the accuracy of trajectory prediction and trajectory planning.

[0119] In some embodiments, the target decoder 412 may be a decoder based on a cross-attention mechanism, see Figure 4 , the target decoder 412 may include a first cross attention unit 4121, a second cross attention unit 4122 and a third cross attention unit 4123; based on the target decoder, the map information feature, the predicted trajectory feature and the historical trajectory feature of the target object are subjected to feature fusion to obtain the first fusion feature, which may include the following steps S11 to S13:

[0120] Step S11, based on the cross attention mechanism, the first predicted trajectory feature and the historical trajectory feature are fused to obtain the second predicted trajectory feature.

[0121] In one implementation, the first predicted trajectory feature can be used as the first query feature (Qurey), and the historical trajectory feature can be used as the first key feature (Key) and the first value feature (Value); the first query feature, the first key feature and the first value feature are input into the first cross-attention unit to generate the second predicted trajectory feature.

[0122] In another implementation, a first query feature can be generated based on the first predicted trajectory feature and the target object relative time dimension interaction feature; a first key feature and a first value feature can be generated based on the historical trajectory feature and the target object relative time dimension interaction feature; the first query feature, the first key feature and the first value feature are input into the first cross-attention unit to generate a second predicted trajectory feature.

[0123] The target object's relative time dimension interaction feature may be an interaction feature for characterizing the target object in the time dimension obtained by embedding and fusion according to the current position information and the historical position information of the target object. Figure 7 As shown, the current position information and historical position information of the target object can be Fourier embedded encoded respectively, and then the encoded vectors are interactively fused based on the pre-trained second network to obtain the target object relative time dimension interactive features. Among them, the second network may include a linear transformation layer (linear). In this way, the interactive influence of the historical moment of the target object on the current moment can be represented by the target object relative time dimension interactive features, so as to improve the accuracy of trajectory planning and trajectory prediction in complex scenarios.

[0124] Step S12: fusing the second predicted trajectory feature and the map information feature based on a cross-attention mechanism to obtain a third predicted trajectory feature.

[0125] In one implementation, the second predicted trajectory feature can be used as the second query feature (Qurey), and the map information feature can be used as the second key feature (Key) and the second value feature (Value); the second query feature, the second key feature and the second value feature are input into the second cross-attention unit to generate a third predicted trajectory feature.

[0126] In another implementation, a second query feature can be generated based on the second predicted trajectory feature and the target object relative position interaction feature; a second key feature and a second value feature can be generated based on the map information feature and the target object relative position interaction feature; the second query feature, the second key feature and the second value feature are input into the second cross-attention unit to generate a third predicted trajectory feature.

[0127] The target object relative position interaction feature may be a feature used to characterize the mutual influence between target objects obtained by embedding and merging the current position information and the historical position information of the target object. Figure 7 As shown, the current position information and the historical position information of the target object can be Fourier embedded encoded respectively, and then the encoded vectors are interactively fused based on the pre-trained third network to obtain the target object relative position interaction feature. The third network may include a linear transformation layer (linear). In this way, the interaction between multiple target objects can be represented by the target object relative position interaction feature to improve the accuracy of trajectory planning and trajectory prediction in complex scenarios.

[0128] Step S13: Fusing the third predicted trajectory feature and the historical trajectory feature based on the cross attention mechanism to obtain a first fused feature.

[0129] In one implementation, the fourth predicted trajectory feature can be used as the third query feature (Qurey), and the map information feature can be used as the third key feature (Key) and the third value feature (Value); the third query feature, the third key feature and the third value feature are input into the third cross-attention unit to generate the fourth predicted trajectory feature.

[0130] In another implementation, a third query feature can be generated according to the third predicted trajectory feature and the target object relative map interaction feature; a third key feature and a third value feature can be generated according to the historical trajectory feature and the target object relative map interaction feature; the third query feature, the third key feature and the third value feature are input into the third cross attention unit to generate the first fusion feature

[0131] The target object relative map interaction feature is a feature obtained by embedding and merging the current position information, historical position information and map element information of the target object and used to characterize the mutual influence between the target object and the map element;

[0132] Continue as Figure 7 As shown, the current position information, historical position information and map element information of the target object can be Fourier embedded encoded respectively, and then the encoded vectors are interactively fused based on the pre-trained fourth network to obtain the target object relative position interaction feature. The fourth network may include a linear transformation layer (linear). In this way, the interaction between the target object and the map element can be represented by the target object relative map interaction feature to improve the accuracy of trajectory planning and trajectory prediction in complex scenarios.

[0133] Using the above method, the target decoder performs feature fusion on the map information features, the predicted trajectory features of the target object and the historical trajectory features based on three cross-attention units connected in sequence. The first fused feature obtained includes the mutual influence among the target vehicle, obstacles and map elements, as well as the interactive influence between the target vehicle and obstacles in the time dimension. Therefore, the first fused feature can be used to form an overall understanding of the dynamic environment, thereby improving the accuracy of trajectory planning and trajectory prediction in complex scenarios.

[0134] Figure 5 is a schematic diagram of the structure of another feature extraction fusion model and trajectory prediction planning model provided by an embodiment of the present disclosure. Figure 5The trajectory prediction planning model 42 may include an ego vehicle trajectory planning module 421 and an obstacle trajectory prediction module 422; the trajectory prediction planning model may generate an initial planning trajectory of the target vehicle and an obstacle prediction trajectory of the obstacle in the following manner: the ego vehicle trajectory planning module may generate an initial planning trajectory of the target vehicle according to the first fusion feature; the obstacle trajectory prediction module may generate an obstacle prediction trajectory of the obstacle according to the first fusion feature, the initial planning trajectory and the current position information of the obstacle.

[0135] In some examples, the first fused feature may be input into an ego vehicle trajectory planning module, and the ego vehicle trajectory planning module may perform trajectory planning according to the first fused feature to generate an initial planned trajectory of the target vehicle.

[0136] Continue to refer Figure 5 The vehicle trajectory planning module 421 may include a first self-attention unit 4211 and a first neural network unit 4212, and the first neural network unit may include a multi-layer perceptron MLP. The first trajectory feature of the target vehicle may be extracted from the first fusion feature based on the first self-attention unit; the first trajectory feature is input into the first neural network unit to obtain the initial planning trajectory of the target vehicle.

[0137] In some examples, the input of the first neural network unit may also include current state information of the target vehicle, such as the current position, speed, acceleration, direction angle, etc. of the target vehicle. In this way, based on the first fusion feature and the current state information of the target vehicle, an initial planning trajectory that meets the requirements of obstacle avoidance, safety, and path smoothness is generated, providing the target vehicle with an optimal driving path to avoid collisions with surrounding obstacles while meeting the requirements of path smoothness and driving efficiency.

[0138] In some examples, the first fused feature, the initial planned trajectory, and the current position information of the obstacle can be input into an obstacle trajectory prediction module, which performs trajectory prediction based on the first fused feature, the initial planned trajectory, and the current position information of the obstacle to generate an obstacle prediction trajectory of the obstacle.

[0139] Continue to refer Figure 5 The obstacle trajectory prediction module 422 may include a fourth cross-attention unit 4221, a second self-attention unit 4222 and a second neural network unit 4223, and the first neural network unit may include a multi-layer perceptron MLP.

[0140] In this example, the first fused feature, the initial planned trajectory and the current position information of the obstacle can be fused based on the fourth cross-attention unit to obtain the second fused feature; the trajectory feature of the obstacle is extracted from the second fused feature based on the second self-attention unit; the trajectory feature is input into the second neural network unit to obtain the obstacle prediction trajectory of the obstacle.

[0141] In one implementation, the first fused feature can be used as the fourth query feature (Query); the initial planning trajectory can be used as the fourth key feature (Key) and the fourth value feature (Value); the fourth query feature, the fourth key feature and the fourth value feature are input into the fourth cross-attention unit to generate the second fused feature.

[0142] In another implementation, the method of fusing the first fused feature, the initial planned trajectory and the current position information of the obstacle based on the fourth cross-attention unit to obtain the second fused feature may include: obtaining the interaction feature of the obstacle relative to the target vehicle after embedding coding and fusing the current position information of the obstacle and the initial planned trajectory; generating a fourth query feature (Query) based on the first fused feature and the interaction feature of the obstacle relative to the target vehicle; generating a fourth key feature (Key) and a fourth value feature (Value) based on the initial planned trajectory and the interaction feature of the obstacle relative to the target vehicle; and inputting the fourth query feature, the fourth key feature and the fourth value feature into the fourth cross-attention unit to generate the second fused feature.

[0143] For example, the current position information of the obstacle and the initial planning trajectory can be Fourier embedded, and then the encoded vectors can be interactively fused based on the pre-trained first network to obtain the interaction features of the obstacle relative to the target vehicle. The first network can include a linear transformation layer. In this way, the fusion extraction of the interaction features between the obstacle and the initial planning trajectory of the target vehicle can be achieved, so as to perform interference analysis and dynamic behavior prediction on the obstacle in combination with the future driving trajectory of the target vehicle, and obtain a more accurate obstacle prediction trajectory.

[0144] In some embodiments of the present disclosure, the target planning trajectory generated by the trajectory planning method in the above-mentioned embodiment can be used to control the vehicle. For example, the vehicle can be controlled to travel according to the generated target planning trajectory. According to the actual vehicle test data, the intersection pass rate is greater than 90%, and the takeover rate is greater than 15 kilometers / time. In particular, for intersections and road-divided-in-two scenarios, the unreasonable lane selection of the vehicle leads to dragon drawing, that is, the vehicle swerves left and right during driving and cannot maintain a straight line. The problem has been significantly improved. The data-driven trajectory planning method can provide a smoother planning trajectory to avoid the dragon drawing problem. In addition, for the abnormal degradation of the autonomous driving function caused by the perception that the lane is too short before the intersection, the missing part can be supplemented based on the above method to avoid abnormal degradation or exit of the autonomous driving function.

[0145] Figure 7 is a flow chart of a trajectory display method provided by an embodiment of the present disclosure. The trajectory display method can be Figure 1 The vehicle and / or server are executed as shown. Figure 7 As shown, the trajectory planning method of this embodiment may include:

[0146] Step S710, determining a target planning trajectory of the target vehicle according to the environmental perception data of the target area.

[0147] Step S720, displaying the target planning trajectory.

[0148] For example, the target planning trajectory can be displayed by a display device of the vehicle and / or server. For example, the target planning trajectory can be displayed on a vehicle-mounted display device at the vehicle end, or the target planning trajectory can be displayed by a device such as a client (such as a mobile phone).

[0149] Among them, the target planning trajectory is a trajectory generated by the first fusion feature output by the feature extraction fusion model, the environmental perception data is used to obtain map element information in the target area and the current position information and historical position information of the target object, the target area is the area including the location of the target vehicle, the target objects include the target vehicle and obstacles, and the feature extraction fusion model is used to output the first fusion feature based on the input current position information, historical position information and map element information.

[0150] The method for generating the target planning trajectory can refer to the description in the aforementioned embodiments of the present disclosure, and will not be repeated here.

[0151] In this way, interactive feedback between target vehicles, obstacles, and map elements can be achieved, efficient and robust trajectory planning can be implemented in a dynamic environment, a more robust target planning trajectory can be generated, and the target planning trajectory can be displayed to the user, making it easier for the user to obtain the future driving direction of the vehicle.

[0152] In some embodiments, the obstacle prediction trajectory and / or relevant information of the map element of the obstacle may also be displayed.

[0153] Figure 8 is a schematic diagram showing a target planning trajectory provided by an embodiment of the present disclosure. Figure 8 As shown in the figure, in a coordinate system (also called a view coordinate system) with the center of the vehicle as the origin and the current driving direction of the vehicle as the x-axis direction, the target area can be exemplarily a rectangular area in the figure, that is, the coordinates of the four corner points of the target area are (100,40), (100,-40), (-40,-40), (-40,40), respectively, and the coordinate unit is meter. As shown in the legend, the target vehicle is the ego vehicle, the blue line is the target planning trajectory of the ego vehicle, such as the target planning trajectory within the last 6 seconds of planning; the red line is the historical trajectory of the ego vehicle, such as the historical trajectory of the ego vehicle within the first 2 seconds; the green line is the historical trajectory of the obstacle, such as the historical trajectory of the obstacle within the first 2 seconds; the brown line is the road boundary line output by the environment perception; the black line is the left and right lane lines output by the environment perception; the azure line is other lines output by the environment perception, such as guide lines, etc. Among them, the guide line can usually appear at an intersection that is too wide, irregular or has complex driving conditions, a ramp of a three-dimensional intersection, or other special locations.

[0154] It should be noted that in this embodiment, one or more of the above lines can be displayed, and other information can also be displayed. For example, only the target planning trajectory of the vehicle can be displayed, or the target planning trajectory and historical trajectory of the vehicle can be displayed, and the obstacle prediction trajectory and other map element related information can also be displayed.

[0155] Fig. 9 Schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. Fig. 9 As shown, the electronic device 1000 may include a memory 1010 and a processor 1020. The memory 1010 may be used to store computer instructions, and the processor 1020 may be used to call computer instructions from the memory 1010 to execute all or part of the steps of any method in the aforementioned embodiments of the present disclosure. The processor may be one or more, and the one or more processors may execute instructions individually or together. The memory may also be one or more, and the one or more memories may store the above-mentioned computer instructions individually or together. Optionally, the electronic device may be Figure 1 Servers and / or vehicles in.

[0156] The embodiment of the present disclosure also provides a vehicle, which may include a memory and a processor, wherein the memory may be used to store computer instructions, and the processor may be used to call computer instructions from the memory to execute all or part of the steps of any method in the aforementioned embodiment of the present disclosure. The processor may be one or more, and the one or more processors may execute instructions individually or collectively, and the memory may also be one or more, and the one or more memories may store the above-mentioned computer instructions individually or collectively.

[0157] The vehicle in the foregoing embodiments of the present disclosure may be an electric vehicle, a hybrid vehicle, a fuel cell vehicle, or other types of vehicles. The vehicle may be an autonomous vehicle or a non-autonomous vehicle. For example, the vehicle provided in this embodiment may be Figure 1 or Figure 2 Vehicle shown.

[0158] The embodiments of the present disclosure also provide a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, any method in the aforementioned embodiments of the present disclosure is implemented. Optionally, the computer-readable storage medium may be a non-transitory storage medium, but is not limited thereto, and may also be a transient storage medium.

[0159] The present disclosure also provides a chip, which may include a processing unit, which may be used to execute all or part of the steps of any method in the foregoing embodiments of the present disclosure. The chip may be a chip in the form of an application-specific integrated chip ASIC, a system on chip SOC, a field programmable gate array FPGA, etc., which is not limited in the present embodiment. Optionally, the chip may also include a storage unit, which may be used to store computer instructions, and the processing unit may be used to call computer instructions from the storage unit to execute all or part of the steps of any method in the foregoing embodiments of the present disclosure.

[0160] The embodiments of the present disclosure further provide a computer program product, which may include a computer program. When the computer program is executed by a processor, any method in the aforementioned embodiments of the present disclosure may be implemented.

[0161] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement any method in the foregoing embodiments of the present disclosure.

[0162] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0163] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0164] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, which may include object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed entirely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be personalized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0165] Various aspects of the present disclosure are described herein with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0166] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so as to produce a machine, so that when these instructions are executed by the processor of the computer or other programmable data processing device, a device for implementing the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions make the computer, programmable data processing device, and / or other equipment work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0167] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0168] The flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to multiple embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of an instruction, and the module, a program segment or a part of an instruction contains one or more executable instructions for implementing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions. It should be noted that implementation by hardware, implementation by software, and implementation by a combination of software and hardware are all equivalent.

[0169] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements to the technology in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein. The scope of the present disclosure is defined by the appended claims.

Claims

1. A trajectory planning method, characterized in that: The method comprises: According to the environmental perception data of the target area, map element information in the target area and current position information and historical position information of the target object are obtained; wherein the target area is an area including the position of the target vehicle whose trajectory is to be planned, and the target object includes the target vehicle and obstacles; Inputting the current location information, the historical location information and the map element information into a feature extraction fusion model to obtain a first fusion feature output by the feature extraction fusion model; wherein the first fusion feature includes an interaction feature characterizing the mutual influence between different target objects and between target objects and map elements; A target planning trajectory of the target vehicle is generated according to the first fusion feature.

2. The method according to claim 1, characterized in that The feature extraction fusion model includes a target encoder and a target decoder; the feature extraction fusion model is used to generate the first fusion feature in the following manner: Based on the target encoder, feature extraction is performed on the map element information, the current position information and the historical position information to obtain map information features, historical trajectory features and first predicted trajectory features of the target object; Based on the target decoder, feature fusion is performed on the map information feature, the historical trajectory feature of the target object and the first predicted trajectory feature to obtain the first fused feature.

3. The method according to claim 2, characterized in that The target decoder is a decoder based on a cross attention mechanism; the target decoder performs feature fusion on the map information feature, the predicted trajectory feature and the historical trajectory feature of the target object to obtain the first fused feature, including: Based on the cross attention mechanism, the first predicted trajectory feature and the historical trajectory feature are integrated to obtain a second predicted trajectory feature; The second predicted trajectory feature and the map information feature are integrated based on a cross attention mechanism to obtain a third predicted trajectory feature; The third predicted trajectory feature and the historical trajectory feature are fused based on a cross attention mechanism to obtain the first fused feature.

4. The method according to claim 3, characterized in that The target decoder includes a first cross attention unit; the first predicted trajectory feature and the historical trajectory feature are fused based on the cross attention mechanism to obtain a second predicted trajectory feature, including: Generate a first query feature according to the first predicted trajectory feature and the target object relative time dimension interaction feature; the target object relative time dimension interaction feature is an interaction feature for characterizing the target object in the time dimension obtained by embedding and fusing the current position information and the historical position information of the target object; Generate a first key feature and a first value feature according to the historical trajectory feature and the target object relative time dimension interaction feature; The first query feature, the first key feature, and the first value feature are input into a first cross attention unit to generate the second predicted trajectory feature.

5. The method according to claim 3, characterized in that: The target decoder further includes a first cross attention unit; the second predicted trajectory feature and the map information feature are fused based on the cross attention mechanism to obtain a third predicted trajectory feature, including: generating a second query feature according to the second predicted trajectory feature and the target object relative position interaction feature; the target object relative position interaction feature is a feature obtained by embedding and fusing the current position information and the historical position information of the target object and used to characterize the mutual influence between the target objects; generating a second key feature and a second value feature according to the map information feature and the target object relative position interaction feature; The second query feature, the second key feature and the second value feature are input into a second cross attention unit to generate the third predicted trajectory feature.

6. The method according to claim 3, characterized in that The fusing the third predicted trajectory feature and the historical trajectory feature based on the cross attention mechanism to obtain the first fused feature includes: generating a third query feature according to the third predicted trajectory feature and the target object relative map interaction feature; the target object relative map interaction feature is a feature obtained by embedding and fusing the current position information, the historical position information and the map element information of the target object and used to characterize the mutual influence between the target object and the map element; generating a third key feature and a third value feature according to the historical trajectory feature and the target object relative map interaction feature; The third query feature, the third key feature and the third value feature are input into a third cross attention unit to generate the first fusion feature.

7. The method according to any one of claims 1 to 6, characterized in that The generating a target planning trajectory of the target vehicle according to the first fusion feature includes: Inputting the first fusion feature into a pre-generated trajectory prediction planning model to obtain an initial planning trajectory of the target vehicle and an obstacle prediction trajectory of the obstacle output by the trajectory prediction planning model; A target planned trajectory of the target vehicle is determined according to the initial planned trajectory and the obstacle predicted trajectory.

8. The method according to claim 7, characterized in that The trajectory prediction planning model includes a vehicle trajectory planning module and an obstacle trajectory prediction module; the trajectory prediction planning model generates the initial planning trajectory of the target vehicle and the obstacle prediction trajectory of the obstacle in the following manner: The ego vehicle trajectory planning module generates an initial planning trajectory of the target vehicle according to the first fusion feature; The obstacle trajectory prediction module generates an obstacle prediction trajectory of the obstacle according to the first fusion feature, the initial planning trajectory and the current position information of the obstacle.

9. The method according to claim 8, characterized in that The self-vehicle trajectory planning module includes a first self-attention unit and a first neural network unit; The vehicle trajectory planning module generates an initial planning trajectory of the target vehicle according to the first fusion feature, including: Extracting a first trajectory feature of the target vehicle from the first fused feature based on a first self-attention unit; The first trajectory feature and the current state information of the target vehicle are input into a first neural network unit to obtain an initial planned trajectory of the target vehicle.

10. The method according to claim 8, characterized in that The obstacle trajectory prediction module includes a fourth cross attention unit, a second self-attention unit and a second neural network unit; the obstacle trajectory prediction module generates an obstacle prediction trajectory of the obstacle according to the first fusion feature, the initial planning trajectory and the current position information of the obstacle, including: fusing the first fused feature, the initial planned trajectory, and the current position information of the obstacle based on a fourth cross attention unit to obtain a second fused feature; Extracting the trajectory feature of the obstacle from the second fused feature based on the second self-attention unit; The trajectory features are input into a second neural network unit to obtain an obstacle prediction trajectory of the obstacle.

11. The method according to claim 10, characterized in that The fusing the first fused feature, the initial planned trajectory and the current position information of the obstacle based on the fourth cross attention unit to obtain a second fused feature includes: After embedding and fusing the current position information of the obstacle and the initial planned trajectory of the target vehicle, the interaction feature of the obstacle relative to the target vehicle is obtained; generating a fourth query feature according to the first fusion feature and the obstacle-to-target-vehicle interaction feature; Generate a fourth key feature and a fourth value feature according to the initial planned trajectory and the obstacle-to-target vehicle interaction feature; The fourth query feature, the fourth key feature and the fourth value feature are input into a fourth cross attention unit to generate the second fusion feature.

12. The method according to claim 7, characterized in that The step of determining the target planned trajectory of the target vehicle according to the initial planned trajectory and the obstacle predicted trajectory includes: Predicting a collision probability between the target vehicle and an obstacle based on the initial planned trajectory and the obstacle predicted trajectory; When the collision probability is greater than or equal to a preset collision probability threshold, updating the initial planned trajectory according to the obstacle prediction trajectory, and using the updated initial planned trajectory as the target planned trajectory to reduce the collision probability; When the collision probability is less than a preset collision probability threshold, the initial planning trajectory is used as the target planning trajectory.

13. A trajectory display method, characterized in that: The method comprises: Determine the target planning trajectory of the target vehicle based on the environmental perception data of the target area; Displaying the target planning trajectory; Among them, the target planning trajectory is a trajectory generated based on the first fusion feature output by the feature extraction fusion model, the environmental perception data is used to obtain the map element information in the target area and the current position information and historical position information of the target object, the target area is the area including the location of the target vehicle, the target object includes the target vehicle and obstacles, and the feature extraction fusion model is used to output the first fusion feature based on the input current position information, the historical position information and the map element information.

14. An electronic device, characterized in that: The invention comprises a memory and a processor, wherein the memory is used to store computer instructions, and the processor is used to call the computer instructions from the memory to execute: the method according to any one of claims 1 to 13.

15. A vehicle, characterized in that: The invention comprises a memory and a processor, wherein the memory is used to store computer instructions, and the processor is used to call the computer instructions from the memory to execute: the method according to any one of claims 1 to 13.

16. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the computer program implements: the method according to any one of claims 1 to 13.

17. A chip, characterized in that: The chip comprises: the chip comprises a processing unit, and the processing unit is used to execute the method as claimed in any one of claims 1 to 13.

Images