Vehicle real-time trajectory prediction method, device and equipment and storage medium

Abstract

The application discloses a vehicle real-time trajectory prediction method, device and equipment and a storage medium, and the method comprises the following steps: performing grid processing on a target road in a first preset time period according to the position of a target vehicle at a current time and the width of a lane where the target vehicle is located, so as to obtain a grid road; wherein the target road is a road where the target vehicle is located; obtaining an obstacle bird's-eye view of the target vehicle in a historical time period and vehicle planning information of the target vehicle in a second preset time period; wherein the obstacle bird's-eye view is a bird's-eye view at the position of an obstacle; the first preset time period belongs to the second preset time period; determining the danger degree of the position of the target vehicle at each time in the second preset time period according to the obstacle bird's-eye view and the vehicle planning information based on a neural network model; and determining the trajectory of the target vehicle at a next time according to the mapping relationship between the danger degree of the position of the target vehicle and the grid road.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, and in particular to a vehicle real-time trajectory method, device, equipment and storage medium. BACKGROUND

[0002] With the development of automatic driving technology, vehicle motion prediction has become a very important part of the vehicle automatic driving system and is one of the foundations of the vehicle automatic driving system.

[0003] However, the existing vehicle trajectory prediction methods, such as the vehicle trajectory prediction method based on physics (vehicle kinematics and dynamics model), the vehicle trajectory prediction method based on traditional machine learning, the vehicle trajectory prediction method based on deep learning and the vehicle trajectory prediction method based on reinforcement learning, give the trajectory of the vehicle in a future period of time by giving a prediction line, which has a complex model and a large time consumption of model training, thereby resulting in poor timeliness of the prediction result and reducing the prediction accuracy. SUMMARY

[0004] The present application provides a vehicle real-time trajectory prediction method, device, equipment and storage medium to simplify the model, reduce the time consumption of model training and improve the prediction accuracy.

[0005] According to an aspect of the present application, a vehicle real-time trajectory prediction method is provided, comprising:

[0006] According to the position of the target vehicle at the current time and the width of the lane where the target vehicle is located, the target road in the first preset time period is rasterized to obtain a rasterized road; wherein the target road is the road where the target vehicle is located;

[0007] Obtain the obstacle bird's eye view of the target vehicle in the historical time period and the vehicle planning information of the target vehicle in the second preset time period; wherein the obstacle bird's eye view is the bird's eye view at the position of the obstacle; the first preset time period belongs to the second preset time period;

[0008] Based on the neural network model, the danger degree of the position of the target vehicle at each time in the second preset time period is determined according to the obstacle bird's eye view and the vehicle planning information;

[0009] According to the mapping relationship between the danger degree of the position of the target vehicle and the rasterized road, the trajectory of the target vehicle at the next time is determined.

[0010] According to another aspect of the present application, a vehicle real-time trajectory prediction device is provided, comprising:

[0011] The grid processing module is configured to perform grid processing on the target road in a first preset time period according to the position of the target vehicle at the current time and the width of the lane where the target vehicle is located, to obtain a grid road; and the target road is the road where the target vehicle is located.

[0012] The data acquisition module is configured to acquire an obstacle bird's-eye view of the target vehicle in a historical time period and vehicle planning information of the target vehicle in a second preset time period; the obstacle bird's-eye view is a bird's-eye view at the position of the obstacle; and the first preset time period belongs to the second preset time period.

[0013] The danger degree determination module is configured to determine, based on a neural network model, a danger degree of the position of the target vehicle at each time in the second preset time period according to the obstacle bird's-eye view and the vehicle planning information.

[0014] The trajectory determination module is configured to determine a trajectory of the target vehicle at a next time according to a mapping relationship between the danger degree of the position of the target vehicle and the grid road.

[0015] According to another aspect of the present application, an electronic device is provided, and the electronic device comprises:

[0016] at least one processor; and

[0017] a memory connected in communication with the at least one processor; wherein

[0018] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the vehicle real-time trajectory prediction method of any one of the embodiments of the present application.

[0019] According to another aspect of the present application, a computer readable storage medium is provided, and the computer readable storage medium stores computer instructions for enabling a processor to implement the vehicle real-time trajectory prediction method of any one of the embodiments of the present application when executed.

[0020] The technical scheme of the embodiment of the present application is that, according to the position of the target vehicle at the current time and the width of the lane where the target vehicle is located, the target road in the first preset time period is rasterized to obtain a rasterized road; the target road is the road where the target vehicle is located; the obstacle bird's eye view of the target vehicle in the historical time period and the vehicle planning information of the target vehicle in the second preset time period are obtained; the obstacle bird's eye view is the bird's eye view at the position of the obstacle; the first preset time period belongs to the second preset time period; based on the neural network model, the danger degree of the position of the target vehicle at each time in the second preset time period is determined according to the obstacle bird's eye view and the vehicle planning information; and the trajectory of the target vehicle at the next time is determined according to the mapping relationship between the danger degree of the position of the target vehicle and the rasterized road. The above technical scheme obtains the rasterized road by rasterizing the target road; the neural network model is used to complete the prediction of the danger degree of each grid in the rasterized road; and then the trajectory of the target vehicle at the next time is determined according to the danger degree of each grid in the rasterized road. The model used in the whole technical scheme is simple, the time consumption of model training is reduced, the trajectory of the target vehicle at the next time can be adjusted in real time according to the changes of external factors, the timeliness of the prediction result is enhanced, and the prediction accuracy is improved.

[0021] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0023] Figure 1A is a flowchart of a vehicle real-time trajectory prediction method provided by the first embodiment of the present application;

[0024] Figure 1B is a schematic diagram of color filling of each grid in the rasterized road according to the first embodiment of the present application;

[0025] Figure 2A is a flowchart of a vehicle real-time trajectory prediction method provided by the second embodiment of the present application;

[0026] Figure 2B is a schematic diagram of establishing a Frenet coordinate system according to the second embodiment of the present application;

[0027] Figure 2C is a schematic diagram of determining a bird's-eye view of a target environment according to an embodiment two of the present application;

[0028] Figure 3 is a structural schematic diagram of a vehicle real-time trajectory prediction device according to an embodiment three of the present application;

[0029] Figure 4 is a structural schematic diagram of an electronic device for implementing a vehicle real-time trajectory prediction method according to an embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to make the personnel in the technical field better understand the present application scheme, the technical scheme in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0031] It should be noted that the terms "target", "first" and "second" and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily have to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0032] In addition, it should also be noted that in the technical scheme of the present application, the collection, storage, use, processing, transmission, provision and disclosure of the position of the target vehicle at the current time, the width of the lane where the target vehicle is located, the obstacle bird's-eye view of the target vehicle in the historical time period and the vehicle planning information of the target vehicle in the second preset time period, etc. The processing conforms to the relevant legal regulations and does not violate public order and good customs.

[0033] Embodiment one

[0034] Figure 1AA flowchart of a vehicle real-time trajectory prediction method provided for Embodiment One of the present application. This embodiment can be applied to the case of predicting the motion trajectory of an autonomous vehicle. The method can be executed by a vehicle real-time trajectory prediction device, which can be realized in the form of hardware and / or software, and can be configured in an electronic device, which can be a vehicle terminal. As shown in FIG. 1, the method comprises the following steps. Figure 1A

[0035] S101, performing grid processing on the target road in the first preset time period according to the position of the target vehicle at the current time and the width of the lane where the target vehicle is located, to obtain a grid road; wherein the target road is the road where the target vehicle is located.

[0036] The target vehicle can refer to an autonomous vehicle that needs to be trajectory predicted. The first preset time period can be set in advance according to actual needs, for example, with the current time as the dividing line, the first preset time period is 5s or 7s in the future. The grid road can refer to the road obtained after the target road is grid processed. The position of the target vehicle at the current time can be obtained through the Global Navigation Satellite System (GNSS) installed on the target vehicle. The width of the lane where the target vehicle is located can be obtained through the RoadSide Unit (RSU) in the vehicle terminal.

[0037] Specifically, taking the position of the target vehicle at the current time as the starting position of the target road sampling, performing grid processing on the target road in the first preset time period according to the width of the lane where the target vehicle is located, and based on the road grid data generation rule, dividing the target road in the first preset time period into grids of the same size or different sizes, to finally obtain the grid road. The road grid data generation rule can be to divide the target road according to the pre-set grid size.

[0038] S102, obtaining the obstacle bird's eye view of the target vehicle in the historical time period and the vehicle planning information of the target vehicle in the second preset time period; wherein the obstacle bird's eye view is the bird's eye view at the position of the obstacle; the first preset time period belongs to the second preset time period.

[0039] ​The historical time period can be pre-set according to actual needs, for example, the current time as a demarcation line, and the historical time period is 1 s or 2 s in the past. The second preset time period is determined according to the first preset time period and the historical time period, for example, the total duration of the second preset time period is equal to the sum of the total duration of the historical time period and the total duration of the first preset time period. For example, taking the current time as a demarcation line, if the first preset time period is 5 s in the future and the historical time period is 1 s in the past, the total duration of the first preset time period is 5 s, the total duration of the historical time period is 1 s, and then the total duration of the second preset time period can be determined as 6 s. It should be noted that the second preset time period includes the first preset time period. The vehicle planning information can include the instantaneous speed and position of the target vehicle at each time in the second preset time period.

[0040] Optionally, the number of the obstacle bird's-eye view is at least 5, so as to ensure that the information of the obstacles around the target vehicle obtained in the historical time period is rich enough, thereby improving the accuracy of determining the risk degree of the position of the target vehicle at each time in the subsequent second preset time period.

[0041] For example, the obstacles around the target vehicle are imaged by four wide-angle cameras installed on the front, back, left and right of the target vehicle; the imaged images are processed and corrected based on preset processing and correction rules to obtain the obstacle bird's-eye view of the target vehicle in the historical time period; and the vehicle planning information of the target vehicle in the second preset time period is obtained from a preset database. The preset database can be pre-set according to actual needs, and the embodiments of the present application do not make specific limitations thereto.

[0042] In S103, based on the neural network model, the risk degree of the position of the target vehicle at each time in the second preset time period is determined according to the obstacle bird's-eye view and the vehicle planning information.

[0043] The neural network model can include a convolutional neural network and a fully connected neural network. The risk degree can refer to the probability of other vehicles in the second preset time period reaching the position of the target vehicle at the same time, in addition to the target vehicle. It should be noted that for each time in the second preset time period, the higher the risk degree of the position of the target vehicle at this time, the higher the possibility of a collision accident between the target vehicle and other vehicles. The other vehicles are other vehicles in addition to the target vehicle.

[0044] Specifically, the obstacle bird's-eye view of the target vehicle in the historical time period and the vehicle planning information of the target vehicle in the second preset time period are taken as inputs of the neural network model, and after processing by the convolutional neural network and the fully connected neural network, the risk degree of the position of the target vehicle at each time in the second preset time period is obtained.

[0045] S104, determining the trajectory of the target vehicle at the next moment according to the mapping relationship between the dangerous degree of the position where the target vehicle is located and the rasterized road.

[0046] The mapping relationship is the corresponding relationship between the dangerous degree of the position where the target vehicle is located at each moment in the first preset time period and each grid in the rasterized road. For each moment in the first preset time period, the dangerous degree of the position where the target vehicle is located at this moment corresponds to one or more grids in the rasterized road at this moment.

[0047] Specifically, the dangerous degree of the position where the target vehicle is located at each moment in the first preset time period is determined according to the dangerous degree of the position where the target vehicle is located at each moment in the second preset time period; and the trajectory of the target vehicle at the next moment is determined according to the mapping relationship between the dangerous degree of the position where the target vehicle is located at each moment in the first preset time period and the rasterized road.

[0048] In an optional embodiment, the dangerous degree of each grid in the rasterized road is determined according to the mapping relationship between the dangerous degree of the position where the target vehicle is located and the rasterized road; each grid in the rasterized road is color-filled according to a preset dangerous degree level; and the trajectory of the target vehicle at the next moment is determined according to the color of each grid in the rasterized road.

[0049] The dangerous degree level can be preset according to experience, and embodiments of the present application do not make specific limitations thereon. For example, it is previously stipulated that a dangerous degree less than or equal to 50% is low risk, a dangerous degree greater than 50% and less than 80% is medium risk, and a dangerous degree greater than or equal to 80% is high risk. It should be noted that for each grid in the rasterized road, the higher the dangerous degree of the grid, the darker the color.

[0050] Specifically, the dangerous degree of each grid in the rasterized road is determined according to the mapping relationship between the dangerous degree of the position where the target vehicle is located at each moment in the first preset time period and the rasterized road; the grids with the same dangerous degree level in the rasterized road are color-filled in the same color according to a preset dangerous degree level; and the trajectory of the target vehicle at the next moment is determined according to the color of each grid in the rasterized road. It should be noted that the color filled in the grids with different dangerous degree levels is different, and the color filled in the grids with the same dangerous degree level is the same.

[0051] For example, referring to FIG. 2, the dangerous degree of the position where the target vehicle is located at each moment in the first preset time period is determined according to the dangerous degree of the position where the target vehicle is located at each moment in the second preset time period. Figure 1BThe white triangle in the figure is the target vehicle, and the white rectangle is other vehicles except the target vehicle. According to the mapping relationship between the dangerous degree of the position of the target vehicle at each time in the first preset time period and the rasterized road, the dangerous degree of each grid in the rasterized road is determined. For each grid in the rasterized road, if the dangerous degree level of the grid is high risk, the color filled in the grid is dark gray; if the dangerous degree level of the grid is medium risk, the color filled in the grid is light gray; if the dangerous degree level of the grid is low risk, the color filled in the grid is white; according to the preset dangerous degree level, the grids with the same dangerous degree level in the rasterized road are filled with the same color; and according to the color of each grid in the rasterized road, one or more grids with the lightest color are selected as the trajectory of the target vehicle at the next time.

[0052] In the optional embodiment, the color of each grid in the rasterized road is filled according to the preset dangerous degree level, so that the driver of the target vehicle can clearly distinguish the dangerous degree of each grid in the rasterized road, and the driver of the target vehicle can quickly determine the trajectory of the target vehicle at the next time according to the color of each grid in the rasterized road.

[0053] The technical scheme of the embodiment of the application comprises the following steps: performing rasterization processing on the target road in the first preset time period according to the position of the target vehicle at the current time and the width of the lane where the target vehicle is located, to obtain a rasterized road; the target road is the road where the target vehicle is located; obtaining an obstacle bird's eye view of the target vehicle in a historical time period and vehicle planning information of the target vehicle in a second preset time period; the obstacle bird's eye view is a bird's eye view of the position of the obstacle; based on a neural network model, determining the dangerous degree of the position of the target vehicle at each time in the second preset time period according to the obstacle bird's eye view and the vehicle planning information; the first preset time period belongs to the second preset time period; and determining the trajectory of the target vehicle at the next time according to the mapping relationship between the dangerous degree of the position of the target vehicle and the rasterized road. The above technical scheme performs rasterization processing on the target road to obtain a rasterized road, and uses a neural network model to complete the prediction of the dangerous degree of each grid in the rasterized road, and then determines the trajectory of the target vehicle at the next time according to the dangerous degree of each grid in the rasterized road. The model used in the whole technical scheme is simple, which reduces the time consumption of model training, and can adjust the trajectory of the target vehicle at the next time in real time according to the changes of external factors, enhances the timeliness of the prediction result, and improves the accuracy of the prediction.

[0054] Embodiment two

[0055] Figure 2AA flowchart of a vehicle real-time trajectory prediction method provided for the second embodiment of the present application, the embodiment is based on the above-mentioned embodiment, and further optimizes the "grid processing of the target road in the first preset time period according to the position of the target vehicle at the current time and the width of the lane where the target vehicle is located", and provides an optional implementation scheme. It should be noted that the parts not described in detail in the embodiment of the present application can refer to the related descriptions of other embodiments. For example Figure 2A The method comprises the following steps:

[0056] S201, establishing a Frenet coordinate system with the midpoint of the rear boundary of the target vehicle as the origin according to the position of the target vehicle at the current time.

[0057] The midpoint of the rear boundary can be a pre-set observation point at the tail of the target vehicle, which is located at the midpoint position of the tail of the target vehicle. The Frenet coordinate system is used to describe the motion state of the target vehicle at each moment.

[0058] Specifically, referring to Figure 2B , the origin of the Frenet coordinate system is determined according to the position of the target vehicle at the current time, the midpoint of the rear boundary of the target vehicle is taken as the origin, and the Frenet coordinate system is established along the s and d directions. It should be noted that Figure 2B The black triangle in the figure is the target vehicle, and the other triangles except the black triangle are other vehicles except the target vehicle; s is the longitudinal coordinate, which represents the distance of the target vehicle driving along the target road; d is the transverse coordinate, which represents the position of the target vehicle left and right on the target road, the positive direction of d represents the position of the target vehicle left on the target road, and the negative direction of d represents the position of the target vehicle right on the target road.

[0059] S202, grid processing of the target road in the first preset time period according to the preset time interval, the instantaneous speed of the target vehicle at each moment in the first preset time period and the width of the lane where the target vehicle is located, to obtain the grid road.

[0060] The preset time interval can be pre-set according to actual needs, such as 0.1s or 0.2s.

[0061] Specifically, the length of each grid in the first preset time period is determined according to the preset time interval and the instantaneous speed of the target vehicle at each moment in the first preset time period; the width of each grid in the first preset time period is determined according to the width of the lane where the target vehicle is located; and then the target road in the first preset time period is grid processed according to the length and width of each grid, to obtain the grid road.

[0062] In an optional embodiment, for each grid, one third of the width of the lane where the target vehicle is located is taken as the width of the grid; the product of the preset time interval and the instantaneous speed of the target vehicle at the current time within the first preset time period is taken as the length of the grid; and the target road within the first preset time period is rasterized according to the length and width of each grid to obtain a rasterized road.

[0063] Specifically, for each grid, if the width of the lane where the target vehicle is located is W, the width of the grid is d1, the length of the grid is s1, the preset time interval is Δt, and the instantaneous speed of the target vehicle at the current time is v, the width and length of the grid are determined by the following formula:

[0064]

[0065] Similarly, the length and width of each grid within the first preset time period can be obtained; and the target road within the first preset time period is rasterized according to the length and width of each grid within the first preset time period to obtain a rasterized road. It should be noted that each grid of the rasterized road has time information, and the specific time of each grid within the first preset time period can be determined. For example, Δt = 0.2s, and the instantaneous speed of the target vehicle at each time within the first preset time period is the same, then the specific time of each i (i is a positive integer) grid within the first preset time period is 0.2 (i-1)s.

[0066] In this optional embodiment, a method for determining the size of each grid in the target road is provided, and then the target road can be rasterized according to the size of each grid in the target road to obtain a rasterized road.

[0067] S203, obtain an obstacle bird's eye view of the target vehicle within a historical time period and vehicle planning information of the target vehicle within a second preset time period; wherein the obstacle bird's eye view is a bird's eye view of the position of the obstacle; and the first preset time period belongs to the second preset time period.

[0068] S204, based on the neural network model, determine the risk degree of the position of the target vehicle at each time within the second preset time period according to the obstacle bird's eye view and the vehicle planning information.

[0069] In an optional embodiment, based on the convolutional neural network in the neural network model, the obstacle bird's eye view, the vehicle planning information, and the bird's eye view of the road around the target vehicle are superimposed to generate a target environment bird's eye view; and based on the fully connected neural network in the neural network model, the risk degree of the position of the target vehicle at each time within the second time period is determined according to the target environment bird's eye view.

[0070] The convolutional neural network is a kind of feedforward neural network containing convolution calculation and having a deep structure, has a representation learning capability, can perform shift-invariant classification on input information according to the hierarchical structure, and is also called a shift-invariant artificial neural network (SIANN). The target environment bird's eye view can refer to a bird's eye view obtained after processing by the convolutional neural network in the neural network model.

[0071] Specifically, the obstacle bird's eye view of the target vehicle in the historical time period, the vehicle planning information of the target vehicle in the second preset time period and the bird's eye view of the road around the target vehicle are input into the convolutional neural network in the neural network model, based on the characteristic that the convolutional neural network can perform shift-invariant classification on input information according to the hierarchical structure, the obstacle bird's eye view of the target vehicle in the historical time period, the vehicle planning information of the target vehicle in the second preset time period and the bird's eye view of the road around the target vehicle are superimposed, referring to Figure 2C , a target environment bird's eye view is generated. It should be noted that, for the sake of illustration, Figure 2C The number of obstacle bird's eye views of the target vehicle in the historical time period in Figure 2C The a in is the obstacle bird's eye view of the target vehicle in the historical time period, the b is the bird's eye view of the road around the target vehicle, and the continuous white lines in the c are used to represent the vehicle planning information of the target vehicle in the second preset time period.

[0072] Further, the target environment bird's eye view is input into the fully connected neural network in the neural network model, and after processing by several layers of fully connected neural network, the danger degree of the position of the target vehicle at each moment in the second time period is obtained.

[0073] In this optional embodiment, a specific method for determining the danger degree of the position of the target vehicle at each moment in the second preset time period is provided, and compared with other deep learning models, the neural network model used in this optional embodiment is relatively simple, reduces the time consumption of model training, and thus can enhance the timeliness of the prediction result and improve the prediction accuracy.

[0074] S205, according to the mapping relationship between the danger degree of the position of the target vehicle and the grid road, the trajectory of the target vehicle at the next moment is determined.

[0075] The technical scheme of the embodiment of the application provides a specific method for determining the danger degree of the position of the target vehicle at each moment in the second preset time period, and perfects the vehicle real-time trajectory prediction method and enhances the applicability of the vehicle real-time trajectory prediction method.

[0076] Embodiment three

[0077] Figure 3 A structural schematic diagram of a vehicle real-time trajectory prediction device is provided for Embodiment Three of the present application. The present embodiment can be applicable to the case of predicting the motion trajectory of an autonomous vehicle. The device can be realized in the form of hardware and / or software and can be configured in an electronic device, which can be a vehicle-mounted terminal. As shown in the figure, the device comprises: Figure 3

[0078] a gridding processing module 301 configured to perform gridding processing on a target road in a first preset time period according to the position of a target vehicle at a current time and the width of a lane in which the target vehicle is located, to obtain a gridded road; wherein the target road is the road in which the target vehicle is located;

[0079] a data acquisition module 302 configured to acquire an obstacle bird's-eye view of the target vehicle in a historical time period and vehicle planning information of the target vehicle in a second preset time period; wherein the obstacle bird's-eye view is a bird's-eye view at the position of an obstacle; the first preset time period belongs to the second preset time period;

[0080] a danger degree determination module 303 configured to determine the danger degree of the position of the target vehicle at each time in the second preset time period based on a neural network model according to the obstacle bird's-eye view and the vehicle planning information;

[0081] a trajectory determination module 304 configured to determine the trajectory of the target vehicle at a next time according to the mapping relationship between the danger degree of the position of the target vehicle and the gridded road.

[0082] The technical solution of the present embodiment determines the trajectory of the target vehicle at a next time by gridding processing the target road in a first preset time period to obtain a gridded road, acquiring an obstacle bird's-eye view of the target vehicle in a historical time period and vehicle planning information of the target vehicle in a second preset time period, determining the danger degree of the position of the target vehicle at each time in the second preset time period, and determining the trajectory of the target vehicle at a next time. The above technical solution gridded processes the target road to obtain a gridded road, uses a neural network model to complete the prediction of the danger degree of each grid in the gridded road, and then determines the trajectory of the target vehicle at a next time according to the danger degree of each grid in the gridded road. The model used in the entire technical solution is simple, which reduces the time consumption of model training, and at the same time, the trajectory of the target vehicle at a next time can be adjusted in real time according to the changes of external factors, thereby enhancing the timeliness of the prediction result and improving the accuracy of the prediction.

[0083] Optionally, the gridding processing module 301 comprises:

[0084] ​The coordinate system establishing unit is configured to establish a Frenet coordinate system with a midpoint of a rear boundary of the target vehicle as an origin according to a position of the target vehicle at a current time;

[0085] The gridding processing unit is configured to perform gridding processing on the target road in the first preset time period according to a preset time interval, an instantaneous speed of the target vehicle at each time in the first preset time period, and a width of a lane in which the target vehicle is located, to obtain a gridded road.

[0086] Optionally, the gridding processing unit is specifically configured to:

[0087] For each grid, one third of the width of the lane in which the target vehicle is located is taken as the width of the grid, and a product of the preset time interval and the instantaneous speed of the target vehicle at the current time in the first preset time period is taken as the length of the grid; the target road in the first preset time period is gridded according to the length and the width of each grid, to obtain the gridded road.

[0088] Optionally, the danger degree determining module 303 is specifically configured to:

[0089] The obstacle bird's-eye view, the vehicle planning information, and the bird's-eye view of the road around the target vehicle are superimposed based on a convolutional neural network in the neural network model, to generate a target environment bird's-eye view; and the danger degree of the position of the target vehicle at each time in the second time period is determined based on a fully connected neural network in the neural network model and the target environment bird's-eye view.

[0090] Optionally, the mapping relationship is a corresponding relationship between the danger degree of the position of the target vehicle at each time in the first preset time period and each grid in the gridded road.

[0091] Correspondingly, the trajectory determining module 304 is specifically configured to:

[0092] The danger degree of each grid in the gridded road is determined according to the mapping relationship between the danger degree of the position of the target vehicle and the gridded road; each grid in the gridded road is color-filled according to a preset danger degree level; and the trajectory of the target vehicle at the next time is determined according to the color of each grid in the gridded road.

[0093] Optionally, the number of the obstacle bird's-eye views is at least 5.

[0094] The vehicle real-time trajectory prediction device provided in the embodiments of the present application can execute the vehicle real-time trajectory prediction method provided in any of the embodiments of the present application, and has the corresponding functional modules and beneficial effects of executing each vehicle real-time trajectory prediction method.

[0095] Embodiment four

[0096] Figure 4 A structural diagram of an electronic device 10 that can be used to implement embodiments of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices (e.g., headsets, glasses, watches, etc.), and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the present application described and / or claimed in this document.

[0097] As shown in Figure 4 The electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., connected to the at least one processor 11 in communication, where the memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer programs stored in the read-only memory (ROM) 12 or loaded into the random access memory (RAM) 13 from the storage unit 18. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0098] A plurality of components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc., an output unit 17, such as various types of displays, speakers, etc., a storage unit 18, such as a magnetic disk, an optical disk, etc., and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunications networks.

[0099] The processor 11 can be various general and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 performs various methods and processes described above, such as the vehicle real-time trajectory prediction method.

[0100] In some embodiments, the vehicle real-time trajectory prediction method can be implemented as a computer program tangibly embodied in a computer readable storage medium, e.g., storage unit 18. In some embodiments, parts or all of the computer program can be loaded and / or installed onto electronic device 10 via, e.g., ROM 12 and / or communication unit 19. When the computer program is loaded onto RAM 13 and executed by processor 11, one or more steps of the vehicle real-time trajectory prediction method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the vehicle real-time trajectory prediction method by other means, e.g., with the aid of firmware.

[0101] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (PLD), a computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0102] Computer programs used to implement the methods of the present application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor, implements the functions / acts specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as part of a standalone software package, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0103] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0104] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0105] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0106] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0107] It should be understood that the various forms of flow shown above can be reordered, added to, or have steps deleted. For example, the steps described in the present application can be performed in parallel, in series, or in a different order, as long as the desired results of the technical solutions of the present application can be achieved, and this is not limited herein.

[0108] The above detailed description does not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A vehicle real-time trajectory prediction method, characterized in that, The method comprises the steps of: According to the position of the target vehicle at the current moment and the width of the lane where the target vehicle is located, the target road in the first preset time period is rasterized to obtain a rasterized road; wherein the target road is the road where the target vehicle is located; Obtain the target vehicle obstacle bird's eye view in the historical time period and the target vehicle planning information in the second preset time period; wherein the obstacle bird's eye view is the bird's eye view at the position of the obstacle; the first preset time period belongs to the second preset time period; Based on the neural network model, according to the obstacle bird's eye view and the vehicle planning information, the risk degree of the target vehicle at each moment in the second preset time period is determined; According to the mapping relationship between the risk degree of the target vehicle position and the rasterized road, the trajectory of the target vehicle at the next moment is determined.

2. The method of claim 1, wherein, According to the position of the target vehicle at the current moment and the width of the lane where the target vehicle is located, the target road in the first preset time period is rasterized to obtain a rasterized road, comprising: According to the position of the target vehicle at the current moment, the midpoint of the rear boundary of the target vehicle is taken as the origin to establish a Frenet coordinate system; According to the preset time interval, the instantaneous speed of the target vehicle at each moment in the first preset time period and the width of the lane where the target vehicle is located, the target road in the first preset time period is rasterized to obtain a rasterized road.

3. The method of claim 2, wherein, According to the preset time interval, the instantaneous speed of the target vehicle at each moment in the first preset time period and the width of the lane where the target vehicle is located, the target road in the first preset time period is rasterized to obtain a rasterized road, comprising: For each grid, one third of the width of the lane where the target vehicle is located is taken as the width of the grid; The product of the preset time interval and the instantaneous speed of the target vehicle at the current moment in the first preset time period is taken as the length of the grid; According to the length and width of each grid, the target road in the first preset time period is rasterized to obtain a rasterized road.

4. The method of claim 1, wherein, The method comprises the steps of: Based on the convolutional neural network in the neural network model, the obstacle bird's eye view, the vehicle planning information and the bird's eye view of the road around the target vehicle are superimposed to generate a target environment bird's eye view; Based on the fully connected neural network in the neural network model, according to the target environment bird's eye view, the risk degree of the target vehicle at each moment in the second preset time period is determined.

5. The method of claim 1, wherein, The mapping relationship is the corresponding relationship between the risk degree of the target vehicle at each moment in the first preset time period and each grid in the rasterized road; Accordingly, the trajectory of the target vehicle at the next moment is determined according to the mapping relationship between the risk degree of the target vehicle position and the rasterized road, comprising: determine, according to a mapping relationship between the dangerous degree of the position of the target vehicle and the rasterized road, the dangerous degree of each grid in the rasterized road; perform color filling on each grid in the rasterized road according to a preset dangerous degree level; determine the trajectory of the target vehicle at the next moment according to the color of each grid in the rasterized road.

6. The method of claim 1, wherein, The number of the obstacle bird's-eye view is at least 5.

7. A vehicle real-time trajectory prediction device characterized by comprising: The method comprises: a rasterization processing module, configured to perform rasterization processing on a target road in a first preset time period according to a position of a target vehicle at a current moment and a width of a lane where the target vehicle is located, to obtain a rasterized road; the target road is a road where the target vehicle is located; a data acquisition module, configured to acquire an obstacle bird's-eye view of the target vehicle in a historical time period and vehicle planning information of the target vehicle in a second preset time period; the obstacle bird's-eye view is a bird's-eye view of an obstacle position; the first preset time period belongs to the second preset time period; a dangerous degree determination module, configured to determine, based on a neural network model, a dangerous degree of a position of the target vehicle at each moment in the second preset time period according to the obstacle bird's-eye view and the vehicle planning information; a trajectory determination module, configured to determine a trajectory of the target vehicle at the next moment according to a mapping relationship between the dangerous degree of the position of the target vehicle and the rasterized road.

8. The apparatus of claim 7, wherein, The rasterization processing module comprises: a coordinate system establishment unit, configured to establish a Frenet coordinate system with a midpoint of a rear boundary of the target vehicle as an origin according to the position of the target vehicle at the current moment; a rasterization processing unit, configured to perform rasterization processing on the target road in the first preset time period according to a preset time interval, an instantaneous speed of the target vehicle at each moment in the first preset time period, and the width of the lane where the target vehicle is located, to obtain the rasterized road.

9. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the vehicle real-time trajectory prediction method in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, and the computer instructions are used to enable the processor to execute the vehicle real-time trajectory prediction method in any one of claims 1-6 when executed.

Images