Vehicle driving intention prediction method, device, electronic equipment and vehicle

Abstract

The present application relates to the field of autonomous driving technology, and in particular to a method, device, electronic device and vehicle for predicting vehicle driving intention. The method comprises: obtaining driving state information of a target vehicle at multiple historical moments; determining the historical intention result corresponding to each historical moment according to the driving state information at each historical moment in multiple historical moments; determining the forgetting factor corresponding to each historical moment; the forgetting factor is used to characterize the degree of influence of the historical intention result at the corresponding historical moment on the target prediction result; the target prediction result is used to characterize the driving intention of the target vehicle; and determining the target prediction result according to the historical intention result and the forgetting factor corresponding to each historical moment in multiple historical moments. The above method reduces the requirements for computer computing power and image clarity, and the analysis and adjustment of its related parameters are more convenient. The use of driving state data and forgetting factors at multiple historical moments can improve the accuracy of vehicle driving intention prediction.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a method, device, electronic device and vehicle for predicting vehicle driving intention. Background Art

[0002] With the rapid development of the current vehicle autonomous driving technology, the safety of autonomous driving technology has always been a difficult problem that needs to be improved. Usually, the autonomous driving vehicle (hereinafter referred to as the self-driving vehicle) plans the driving trajectory of the self-driving vehicle in a certain period of time in the future based on the driving intentions (such as lane change intentions) of the target vehicles around it, which can effectively improve the safety performance of the self-driving vehicle. For example, based on the lane change intention of the target vehicle cutting in front of the self-driving vehicle, the self-driving vehicle plans the driving trajectory in the future in advance, and adjusts the movement decision of the self-driving vehicle in time to avoid collision with the target vehicle. However, in the complex outdoor environment, the driving intention of the target vehicle is highly uncertain.

[0003] The accuracy of the driving intention obtained based on the deep learning model is relatively high, and the deep learning model is usually used in the prior art to predict the intention of the target vehicle. However, on the one hand, the deep learning model usually includes hundreds of thousands to tens of millions of parameters, so that the training of the deep learning model requires high computer computing power; on the other hand, the deep learning model is usually obtained based on the image RGB convolutional neural network, and it is necessary to extract the feature information (such as lane line information) required for the prediction of the target vehicle's driving intention from the high-definition map image. In addition, since the deep learning model requires high computer computing power and high-definition map images, the cost of predicting the target vehicle's driving intention is relatively high. In addition, since it is difficult to conduct detailed analysis and accurately adjust the parameters of the deep learning model for specific scenarios, the subsequent optimization of the deep learning model is unfavorable. For example, it takes a lot of time and cost to retrain the deep learning model for information of specific scenarios.

[0004] Therefore, it is necessary to provide a vehicle driving intention prediction method, device, electronic equipment and vehicle to address the problems of high computer computing power consumption, high image clarity requirements, high cost, and difficulty in analyzing and adjusting parameters in the process of predicting the target vehicle's driving intention. Summary of the invention

[0005] The embodiment of the present application provides a method, device, electronic device and vehicle for predicting the driving intention of a vehicle. The method determines the historical intention results corresponding to each historical moment based on the driving status information of the target vehicle at multiple historical moments, and sets the degree of influence of each historical intention result on the target prediction result through a forgetting factor, wherein the target prediction result is used to characterize the driving intention of the target vehicle. On the one hand, the above method does not need to be based on a deep learning model, and does not need to use high-definition image information, which reduces the requirements for computer computing power and image clarity, and saves the cost of predicting the driving intention of the target vehicle. On the other hand, the driving status data of multiple historical moments is used to further improve the accuracy of the vehicle driving intention prediction. In addition, the weight corresponding to the corresponding historical intention result at each historical moment is controlled by the forgetting factor to ensure the prediction effect of the target vehicle's driving intention. In addition, the analysis and adjustment of the relevant parameters of the mathematical model obtained based on the above method are more convenient.

[0006] The present application is introduced below from multiple aspects, and the implementation methods and beneficial effects of the following multiple aspects can be referenced to each other.

[0007] In a first aspect, an embodiment of the present application provides a method for predicting a vehicle's driving intention, the method comprising:

[0008] Obtain the driving status information of the target vehicle at multiple historical moments;

[0009] Determine, according to the driving state information of each historical moment in the plurality of historical moments, a historical intention result corresponding to each historical moment;

[0010] Determine the forgetting factor corresponding to each historical moment; the forgetting factor is used to characterize the influence of the historical intention result of the corresponding historical moment on the target prediction result; the target prediction result is used to characterize the driving intention of the target vehicle;

[0011] A target prediction result is determined according to the historical intention result and the forgetting factor corresponding to each historical moment in multiple historical moments.

[0012] In an implementation of the first aspect, the determining of the forgetting factor corresponding to each historical moment includes:

[0013] The forgetting factor corresponding to each historical moment is determined in chronological order; the degree of influence represented by each forgetting factor increases in chronological order.

[0014] The above method further improves the accuracy of vehicle driving intention prediction by setting the weights of the historical intention costs corresponding to each historical moment (i.e. the above forgetting factor) to increase in chronological order.

[0015] In one implementation of the first aspect above, the driving state information at each historical moment includes the lateral speed of the target vehicle, the heading angle of the target vehicle, the vehicle width of the target vehicle, the lane width of the current lane where the target vehicle is located, and the lateral position of the target vehicle relative to the current lane;

[0016] The above determination of the historical intention results corresponding to each historical moment includes:

[0017] Determine a lateral offset ratio based on the vehicle width, lane width and lateral position corresponding to each historical moment;

[0018] Based on the lateral deviation ratio, lateral velocity and heading angle, the historical intention result corresponding to each historical moment is determined.

[0019] In an implementation of the first aspect, the method further includes:

[0020] Obtaining the interactive information between the target vehicle and the surrounding vehicles of the target vehicle at the current moment; the surrounding vehicles include vehicles within a preset distance from the target vehicle;

[0021] Determine a first auxiliary driving intention result according to the interactive information;

[0022] The driving intention of the target vehicle is output according to the first auxiliary driving intention result and the target prediction result.

[0023] In an implementation of the first aspect, the interaction information includes at least one of momentum-related information of the surrounding vehicle, the longitudinal velocity of the surrounding vehicle, the longitudinal spacing between the surrounding vehicle and the target vehicle, the target longitudinal velocity of the target vehicle and the target longitudinal acceleration of the target vehicle.

[0024] Since the speed relationship between the surrounding vehicle and the target vehicle, the longitudinal distance between the surrounding vehicle and the target vehicle, the target longitudinal acceleration of the target vehicle and the momentum of the surrounding vehicle will significantly affect the target vehicle's intention, considering the interaction scenario between the target vehicle and its surrounding vehicles during the prediction of the target vehicle's driving intention is conducive to improving the accuracy of the target vehicle's driving intention prediction.

[0025] In an implementation of the first aspect, the method further includes:

[0026] Obtaining traffic information related to the target vehicle; the traffic information includes lane line type and / or traffic light information;

[0027] Determine the second auxiliary driving intention result corresponding to the traffic information according to the preset traffic rules;

[0028] The driving intention of the target vehicle is output according to the second auxiliary driving intention result and the target prediction result.

[0029] In an implementation of the first aspect, the method further includes:

[0030] Acquire current weather information of the area where the target vehicle is located; the current weather information includes at least one of rainfall and weather type;

[0031] determining a third auxiliary driving intention result according to current weather information;

[0032] The driving intention of the target vehicle is output according to the third auxiliary driving intention result and the target prediction result.

[0033] In an implementation of the first aspect, the driving intention includes at least one of an intention to change lanes to the left, an intention to keep lane, and an intention to change lanes to the right.

[0034] In a second aspect, an embodiment of the present application provides a vehicle driving intention prediction device, the device comprising:

[0035] A state acquisition module is used to obtain the driving state information of the target vehicle at multiple historical moments;

[0036] A state determination module, used to determine a historical intention result corresponding to each historical moment according to the driving state information of each historical moment among the multiple historical moments;

[0037] A forgetting factor determination module is used to determine the forgetting factor corresponding to each historical moment; the forgetting factor is used to characterize the influence of the historical intention result of the corresponding historical moment on the target prediction result; the target prediction result is used to characterize the driving intention of the target vehicle;

[0038] The intention determination module is used to determine the target prediction result according to the historical intention result and the forgetting factor corresponding to each historical moment in multiple historical moments.

[0039] In an implementation of the second aspect, the forgetting factor determination module includes:

[0040] A first forgetting factor determination submodule is used to determine the time sequence of multiple historical moments;

[0041] The second forgetting factor determination submodule is used to determine the forgetting factor corresponding to each historical moment; wherein the influence degree represented by each forgetting factor increases in chronological order.

[0042] In an implementation of the second aspect, the driving state information at each historical moment includes the lateral speed of the target vehicle, the heading angle of the target vehicle, the vehicle width of the target vehicle, the lane width of the current lane where the target vehicle is located, and the lateral position of the target vehicle relative to the current lane;

[0043] The above intention determination module includes:

[0044] A first intention determination submodule, for determining a lateral offset ratio based on a vehicle width, a lane width, and a lateral position corresponding to each historical moment;

[0045] The second intention determination submodule is used to determine the historical intention result corresponding to each historical moment based on the lateral offset ratio, the lateral speed and the heading angle.

[0046] In an implementation of the second aspect, the apparatus further includes:

[0047] An interaction acquisition module is used to acquire the interaction information between the target vehicle and the surrounding vehicles of the target vehicle at the current moment; the surrounding vehicles include vehicles within a preset distance from the target vehicle;

[0048] An interaction determination module, used to determine a first auxiliary driving intention result according to the interaction information;

[0049] The interactive output module is used to output the driving intention of the target vehicle according to the first auxiliary driving intention result and the target prediction result.

[0050] In an implementation of the second aspect, the interactive information includes at least one of momentum-related information of the surrounding vehicle, the longitudinal velocity of the surrounding vehicle, the longitudinal spacing between the surrounding vehicle and the target vehicle, the target longitudinal velocity of the target vehicle and the target longitudinal acceleration of the target vehicle.

[0051] In an implementation of the second aspect, the apparatus further includes:

[0052] A traffic acquisition module is used to acquire traffic information related to the target vehicle; the traffic information includes lane line type and / or traffic light information;

[0053] A traffic determination module, used to determine a second auxiliary driving intention result corresponding to the traffic information according to preset traffic rules;

[0054] The traffic output module is used to output the driving intention of the target vehicle according to the second auxiliary driving intention result and the target prediction result.

[0055] In an implementation of the second aspect, the apparatus further includes:

[0056] A weather acquisition module, used to acquire current weather information of the area where the target vehicle is located; the current weather information includes at least one of rainfall and weather type;

[0057] A weather determination module, used to determine a third auxiliary driving intention result according to current weather information;

[0058] The weather output module is used to output the driving intention of the target vehicle according to the third auxiliary driving intention result and the target prediction result.

[0059] In an implementation of the second aspect, the driving intention includes at least one of an intention to change lanes to the left, an intention to keep lane, and an intention to change lanes to the right.

[0060] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the above-mentioned vehicle driving intention prediction method.

[0061] In a fourth aspect, an embodiment of the present application provides a vehicle, which includes the electronic device of the third aspect.

[0062] The embodiment of the present application provides a method, device, electronic device and vehicle for predicting the driving intention of a vehicle. The method determines the historical intention results corresponding to each historical moment based on the driving state information of the target vehicle at multiple historical moments, and sets the influence degree of each historical intention result on the target prediction result through the forgetting factor, wherein the target prediction result is used to characterize the driving intention of the target vehicle. On the one hand, the above method does not need to be based on a deep learning model, and does not need to use high-definition image information, which reduces the requirements for computer computing power and image clarity, and saves the cost of predicting the driving intention of the target vehicle. On the other hand, the driving state data of multiple historical moments are used to avoid inaccurate prediction results based on one moment, and the accuracy of the target prediction result has a strong dependence on the data accuracy of the perception unit and the fusion unit. In addition, the weight corresponding to the corresponding historical intention results of each historical moment is controlled by the forgetting factor, so as to further improve the accuracy of the vehicle driving intention prediction. In addition, the mathematical model obtained based on the above method can be specifically analyzed for a specific scene, so as to analyze and adjust the parameters of the mathematical model to achieve the purpose of optimizing the mathematical model. For example, according to different scenes, the forgetting factor is adjusted to ensure the vehicle driving intention prediction effect in different scenes and improve the accuracy of the vehicle driving intention prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0064] Figure 1 is a scene diagram of a vehicle driving intention prediction method provided in an embodiment of the present application;

[0065] Figure 2 is a schematic diagram of an automatic driving control method based on vehicle driving intention prediction provided in an embodiment of the present application;

[0066] Figure 3 It is a flow chart of a method for predicting vehicle driving intention provided in an embodiment of the present application;

[0067] Figure 4A It is a flow chart of a method for predicting vehicle driving intention based on an interactive scenario provided in an embodiment of the present application;

[0068] Figure 4B It is a flow chart of a method for predicting vehicle driving intention based on traffic scenarios provided in an embodiment of the present application;

[0069] Figure 4C It is a flow chart of a vehicle driving intention prediction method based on weather scenarios provided in an embodiment of the present application;

[0070] Figure 5 It is a structural schematic diagram of a vehicle driving intention prediction device provided in an embodiment of the present application;

[0071] Figure 6 It is a block diagram of an electronic device for a vehicle driving intention prediction method provided in an embodiment of the present application. DETAILED DESCRIPTION

[0072] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0073] "One embodiment" or "embodiment" referred to herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present invention. In the description of the present invention, it should be understood that the terms "first", "second", "third" and "fourth" in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units that are inherent to these processes, methods, products or devices.

[0074] Before introducing the vehicle driving intention prediction method of the embodiment of the present application, the scenario of the vehicle driving intention prediction method is first introduced.

[0075] See also Figure 1 , Figure 1 It is a scene diagram of a vehicle driving intention prediction method provided by an embodiment of the present application, in which a target vehicle 102 is within a preset distance range in front of the self-vehicle 101 (in the driving direction), and the surrounding vehicles of the target vehicle 102 (referred to as surrounding vehicles) include: the first vehicle directly in front of the target vehicle 102 - the front vehicle 1021, the first vehicle in front of the left lane of the target vehicle 102 - the left front vehicle 1022, the first vehicle behind the left lane of the target vehicle 102 - the left rear vehicle 1023, the first vehicle in front of the right lane of the target vehicle 102 - the right front vehicle 1024, and the first vehicle behind the right lane of the target vehicle 102 - the right rear vehicle 1025.

[0076] During the driving process of the vehicle 101, the future driving trajectory of the target vehicle 102 will affect the driving state of the vehicle 101. For example, the trajectory of the target vehicle 102 changing lanes to the right may affect the driving speed of the vehicle 101. The driving trajectory is determined by the driving intention of the target vehicle 102 at the current moment (such as the intention to change lanes to the right). Therefore, the accurate prediction of the driving intention of the target vehicle 102 by the vehicle 101 can enable the vehicle 101 to adjust the driving state information such as the vehicle speed and heading angle in time to avoid accidents such as collisions with the target vehicle 102.

[0077] See also Figure 2 , Figure 2 1 is a schematic diagram of an automatic driving control method based on target vehicle driving intention prediction provided by an embodiment of the present application. The driving control system of the vehicle 101 includes: a navigation and map positioning unit, a perception unit, a body unit, a fusion unit, a target selection unit, a prediction unit, a decision unit and an execution unit.

[0078] The navigation and map positioning unit, the perception unit and the body unit are used to generate data required by the control method. The navigation and map positioning unit is used to obtain global path information, road information, positioning information, lane line information, weather information, etc. The navigation and map positioning unit can be implemented through a high-precision map; the perception unit is used to obtain target vehicle information (such as the target vehicle 102), traffic lights and lane line information, etc. The perception unit can be implemented through a camera, radar, laser radar, etc.; the body unit is used to obtain driver request, settings and vehicle speed information, such as the vehicle speed information, acceleration information, steering information, and turn signal information of the vehicle 101.

[0079] The fusion unit is used to obtain and fuse the data required by the above control method from the navigation and map positioning unit, the perception unit and the body unit to obtain global path information including route selection information, lane information of the ego vehicle, lane change priority request of the ego vehicle, etc. The target selection unit is used to select target vehicles within a preset distance range.

[0080] The prediction unit is used to predict the driving intention and trajectory of the target vehicle, wherein the trajectory of the target vehicle is predicted based on the predicted driving intention. The prediction unit obtains the data required for the prediction from the navigation and map positioning unit, the fusion unit and the target selection unit. For example, the prediction unit obtains global path information, road information, positioning information, lane line information, weather information and other data from the navigation and map positioning unit, the prediction unit obtains information such as traffic lights from the fusion unit, and the prediction unit obtains the location information of the target vehicle from the target selection unit.

[0081] The decision unit is used to process the information of the fusion unit and the prediction unit, and determine the request instructions such as the steering wheel angle request, acceleration and deceleration request, and turn signal request. The execution unit is used to execute the above request instructions sent by the decision unit. The parameter transmission between the above units includes but is not limited to the transmission of CAN bus and Ethernet.

[0082] The prediction unit includes an intention prediction model based on preset conditions, and the intention prediction model is used to predict the intention of the target vehicle. For example, the intention is predicted for the target vehicle 102. In addition, it can be understood that the driving control system is jointly determined by a variety of units and electronic devices, and may include other units other than the above-mentioned units, which are not limited here. The driving control system includes but is not limited to devices and electronic devices for implementing the above-mentioned driving control method. The above-mentioned fusion unit, target selection unit, prediction unit and decision unit can be set in the vehicle 101, or can be set in a cloud server or other equipment that is not set in the vehicle 101, which is not limited here.

[0083] As mentioned above, the existing target vehicle driving intention prediction process uses deep learning models, which results in problems such as wasted computer computing power, high image clarity requirements, high costs, and difficulty in analyzing and adjusting parameters.

[0084] The present application determines the historical intention results corresponding to each historical moment based on the driving state information of the target vehicle at multiple historical moments, and sets the influence degree of each historical intention result on the target prediction result through the forgetting factor, wherein the target prediction result is used to characterize the driving intention of the target vehicle. On the one hand, the above method does not need to be based on a deep learning model, and does not need to use high-definition image information, which reduces the requirements for computer computing power and image clarity, and saves the cost of predicting the driving intention of the target vehicle. On the other hand, the driving state data of multiple historical moments are used to avoid predicting the driving trajectory based on the driving state data at one moment, which causes inaccurate target prediction results and strong dependence of the target prediction results on the accuracy of the driving state data at that moment, thereby improving the accuracy of the target prediction results. In addition, the weight corresponding to the corresponding historical intention results at each historical moment is controlled by the forgetting factor, so as to further improve the accuracy of the vehicle driving intention prediction. In addition, the mathematical model obtained based on the above method can be specifically analyzed for specific scenarios, so as to analyze and adjust the parameters of the mathematical model to achieve the purpose of optimizing the mathematical model. For example, according to different scenarios, the forgetting factor is adjusted to ensure the prediction effect of the vehicle driving intention in different scenarios.

[0085] The following describes a specific embodiment of a vehicle driving intention prediction method of the present application. Figure 3 It is a flow chart of a vehicle driving intention prediction method provided in an embodiment of the present application. This specification provides method operation steps such as the embodiment or flow chart, but may include more or fewer operation steps based on conventional or non-creative labor. The order of steps listed in the embodiment is only one way of executing the steps among many orders, and does not represent the only order of execution. When the actual system or server product is executed, it can be executed in sequence or in parallel according to the method shown in the embodiment or the accompanying drawings (for example, in a parallel processor or multi-threaded processing environment). The following is combined with Figure 1 The scene shown is Figure 3 The method shown is described in detail, as shown in Figure 3 As shown, the method includes:

[0086] S301: Obtain driving status information of a target vehicle at multiple historical moments.

[0087] In some embodiments of the present application, the driving status information includes the lateral speed of the target vehicle, the heading angle of the target vehicle, the vehicle width of the target vehicle, the lane width of the current lane where the target vehicle is located, and the lateral position of the target vehicle relative to the current lane.

[0088] S303: Determine the historical intention result corresponding to each historical moment according to the driving status information of each historical moment among the multiple historical moments.

[0089] In some embodiments of the present application, determining the historical intention result corresponding to each historical moment specifically includes: determining the lateral offset ratio based on the vehicle width, lane width and lateral position corresponding to each historical moment; determining the historical intention result corresponding to each historical moment based on the lateral offset ratio, lateral speed and heading angle.

[0090] S305: Determine the forgetting factor corresponding to each historical moment.

[0091] The longer the time interval between each historical moment in multiple historical moments and the current moment, the lower the impact of the historical intention cost corresponding to the historical moment on the target intention result.

[0092] In some embodiments of the present application, the forgetting factor corresponding to each historical moment is determined in chronological order; the influence degree represented by each forgetting factor increases in chronological order. By setting the weight of the historical intention cost corresponding to each historical moment (i.e., the above-mentioned forgetting factor) to increase in chronological order, the accuracy of the vehicle driving intention prediction is improved.

[0093] S307: Determine a target prediction result according to the historical intention result and the forgetting factor corresponding to each historical moment in the multiple historical moments.

[0094] For example, according to Figure 3 The mathematical model obtained by the method shown includes lateral costs and straight costs determined according to state restriction rules; the lateral costs and straight costs use the genetic factor λ to achieve the weight of the historical intention results corresponding to each historical moment increasing in chronological order.

[0095] The above horizontal costs are calculated using the following formulas (1) and (2):

[0096]

[0097]

[0098] in, is the horizontal cost, is the lateral displacement ratio at each historical moment, θ 1 ···θ N is the heading angle at each historical moment, is the lateral velocity at each historical moment, λ is a real number between 0 and 1, k d-R , k d-theta , are the coefficients of each item; R lat is the lateral offset ratio, P lat is the horizontal position, w obs is the vehicle width, w lane is the lane width.

[0099] The above straight travel cost is calculated by the following formula (3):

[0100]

[0101]

[0102] in, is the straight travel cost, k R_chg , k theta-chg , k v_chg are the coefficients of each item. and It is used to characterize the historical intention cost of the Nth historical moment. The characterization method of the historical intention cost of other historical moments is similar to that of the Nth historical moment and will not be elaborated here.

[0103] As mentioned above, in the prior art, due to the use of deep learning models to predict vehicle driving intentions, the parameters of the deep learning models are very large. It is difficult to analyze and adjust the parameters of the deep model specifically for a specific scenario. The mathematical model obtained by the above method in the present application can improve the accuracy of vehicle driving intention prediction by analyzing and adjusting the specific parameters of the mathematical model for a specific scenario. The method of analyzing and adjusting the specific parameters of the mathematical model for a specific scenario is described in detail below, wherein the specific scenario takes the interaction between the target vehicle and the surrounding vehicles (surrounding vehicles) of the target vehicle, traffic and weather as an example, but the specific scenario of the present application is not limited to this.

[0104] Regarding the interaction between the target vehicle and its surrounding vehicles. The speed relationship between the surrounding vehicle and the target vehicle, the longitudinal distance between the surrounding vehicle and the target vehicle, the target longitudinal acceleration of the target vehicle, and the momentum of the surrounding vehicle will significantly affect the target vehicle's driving intention. For example, when the momentum of the surrounding vehicle is large, the mechanical reaction speed of the surrounding vehicle to avoid other vehicles is slow (such as slow speed change), and the target vehicle usually keeps a distance from the surrounding vehicle in order to respond to the driving state changes of the surrounding vehicle in time and avoid collision. Therefore, in the process of predicting the target vehicle's driving intention, considering the interaction scenario between the target vehicle and its surrounding vehicles is conducive to improving the accuracy of the target vehicle's driving intention prediction.

[0105] Figure 4A This is a flow chart of a method for predicting vehicle driving intention based on an interactive scenario provided in an embodiment of the present application. The following is a detailed description of the method for predicting vehicle driving intention based on an interactive scenario between a target vehicle and its surrounding vehicles.

[0106] S309A: Obtaining the interaction information between the target vehicle and surrounding vehicles of the target vehicle at the current moment. The surrounding vehicles include vehicles within a preset distance from the target vehicle.

[0107] In some embodiments of the present application, the above-mentioned interactive information includes at least one of momentum-related information of the surrounding vehicle, the longitudinal velocity of the surrounding vehicle, the longitudinal distance between the surrounding vehicle and the target vehicle, the target longitudinal velocity of the target vehicle and the target longitudinal acceleration of the target vehicle.

[0108] In some embodiments of the present application, the surrounding vehicles include the front vehicle, the left front vehicle, the left rear vehicle, the right front vehicle, and the right rear vehicle. The front vehicle is the first vehicle in front of the target vehicle in the current lane (e.g. Figure 1 The target vehicle 102 and the front vehicle 1021 shown in the figure, the left front vehicle and the left rear vehicle are respectively the first vehicle in front of the target vehicle and the first vehicle behind the target vehicle on the left lane of the target vehicle (such as Figure 1 The left front vehicle 1022 and the left rear vehicle 1023 shown in the figure), the right front vehicle and the right rear vehicle are respectively the first vehicle in front of the target vehicle and the first vehicle behind the target vehicle on the right lane of the target vehicle (such as Figure 1 The right front vehicle 1024 and the right rear vehicle 1025 are shown).

[0109] In some embodiments of the present application, the interactive information includes the longitudinal speed of the surrounding vehicle, the longitudinal distance between the surrounding vehicle and the target vehicle, the volumes of the left front vehicle, the left rear vehicle, the right front vehicle and the right rear vehicle, and the target longitudinal speed and target longitudinal acceleration of the target vehicle.

[0110] S311A: Determine a first auxiliary driving intention result according to the interactive information.

[0111] S313A: Outputting the driving intention of the target vehicle according to the first auxiliary driving intention result and the target prediction result.

[0112] For example, the mathematical model also includes the left interaction cost, the right interaction cost and the longitudinal interaction cost determined according to the interaction restriction rules of the surrounding vehicles on the driving intention. The first auxiliary driving intention result is characterized based on the left interaction cost, the right interaction cost and the longitudinal interaction cost.

[0113] The above left-side interaction cost is calculated using formula (4):

[0114] C L =k LF *d LF +k ep-LF *v LF *V LF +k ego-LF *(v LF -v S )+k LR *d LR +k ep-LR *

[0115] v LR*V LR +k ego-L R*(ν S -v LR ) (4)

[0116] Among them, C L is the left-side interaction cost, ν LF is the longitudinal velocity of the left front vehicle, V LF is the volume of the left front car, ν LR is the longitudinal velocity of the left rear vehicle, V LR is the volume of the left rear vehicle, d LF is the longitudinal distance between the target vehicle and the left front vehicle, d LR is the longitudinal distance between the target vehicle and the left rear vehicle.

[0117] The above right-side interaction cost is calculated using formula (5):

[0118] C R =k RF *d RF +k ep-RF *v RF *V RF +k ego-RF *(v RF -v S )+k RR *d RR +k ep-R R*ν RR *

[0119] V RR +k ego-R R*(v S -ν RR ) (5)

[0120] Among them, C R is the right-hand interaction cost, ν RF is the longitudinal velocity of the right front vehicle, V RF is the volume of the right front car, v RR is the longitudinal velocity of the right rear vehicle, V RR is the volume of the right rear car, f RF is the longitudinal distance between the target vehicle and the right front vehicle, d RR is the longitudinal distance between the target vehicle and the right rear vehicle.

[0121] The above vertical interaction cost is calculated using formula (6):

[0122] C eff =k s-aclr *a s +k s-dist *d fwd +kego-fwd *(v fwd -v s ) (6)

[0123] Among them, C eff is the vertical interaction cost, a s is the target longitudinal acceleration, d fwd is the longitudinal distance between the target vehicle and the vehicle in front, v fwd is the longitudinal velocity of the vehicle in front, v s is the target longitudinal velocity, k s-aclr , k s-dist , k ego-fwd is the coefficient.

[0124] For the traffic scenes related to the target vehicle. Since the vehicle must abide by traffic rules during driving to ensure the normal operation of traffic and avoid traffic accidents, therefore, in the process of predicting the target vehicle's driving intention, considering the traffic scenes related to the target vehicle is conducive to improving the accuracy of the target vehicle's driving intention prediction.

[0125] Figure 4B This is a flow chart of a method for predicting vehicle driving intention based on traffic scenarios provided in an embodiment of the present application. The method for predicting vehicle driving intention based on traffic scenarios related to target vehicles is described in detail below.

[0126] S309B: Obtaining traffic information related to the target vehicle, wherein the traffic information includes lane line type and / or traffic light information.

[0127] S311B: Determine a second auxiliary driving intention result corresponding to the traffic information according to preset traffic rules.

[0128] For example, if the traffic information includes that the indicator light ahead is red, then the second auxiliary driving intention result is determined to be that going straight and overtaking the road is prohibited.

[0129] S313B: Output the driving intention of the target vehicle according to the second auxiliary driving intention result and the target prediction result.

[0130] For example, the mathematical model further includes a traffic cost determined according to the traffic restriction rules of the traffic rules for the driving intention, wherein the traffic cost is used to characterize the second auxiliary driving intention result.

[0131] The above transportation costs are calculated using formulas (7), (8) and (9):

[0132] C Reg =k ln *C Ln +k lt *C Lt (7)

[0133]

[0134]

[0135] Among them, C Reg is the transportation cost, C Ln is the lane cost associated with the lane type, C Lt is the traffic light cost associated with the traffic light condition, k ln , k lt is the coefficient, b sld , b r , b g is a constant term.

[0136] For the current weather scene in the area where the target vehicle is located. Bad weather has a greater impact on the vehicle's intention to change lanes. In bad weather, lane lines are difficult to identify, and the vehicle's field of vision is poor, which reduces the probability of the vehicle's intention to change lanes. Therefore, when predicting the target vehicle's driving intention, considering the current weather scene in the area where the target vehicle is located is conducive to further improving the accuracy of the target vehicle's driving intention prediction.

[0137] Figure 4C This is a flow chart of a method for predicting vehicle driving intention based on weather scenarios provided in an embodiment of the present application. The method for predicting vehicle driving intention based on weather scenarios is described in detail below.

[0138] S309C: Obtain current weather information of the area where the target vehicle is located. The weather information includes at least one of rainfall and weather type.

[0139] S311C: Determine a third auxiliary driving intention result according to current weather information.

[0140] For example, if the current weather is foggy, the third auxiliary driving intention result is determined to be prohibiting lane change.

[0141] S313C: Output the driving intention of the target vehicle according to the third auxiliary driving intention result and the target prediction result.

[0142] For example, the mathematical model further includes a weather cost determined according to weather restriction rules for driving intentions based on weather conditions; the weather cost is used to characterize the third auxiliary driving intention result.

[0143] The above weather cost is calculated using the following formula (10):

[0144]

[0145] Among them, C we is the weather cost, Rr is the rainfall, k r is the coefficient, b Ex is a constant term. In extreme weather conditions, the output constant term b Ex .

[0146] At the same time, in some embodiments of the present application, the driving intention of the target vehicle is output based on the above-mentioned target prediction result and at least one of the above-mentioned first auxiliary driving intention result, the second auxiliary driving intention result and the third auxiliary driving intention result.

[0147] In some embodiments of the present application, the above-mentioned driving intention includes at least one of an intention to change lanes to the left, an intention to keep the lane, and an intention to change lanes to the right.

[0148] In some embodiments of the present application, the mathematical model further includes intention costs of intention to change lane to the left, intention to keep lane, and intention to change lane to the right.

[0149] The intention cost is calculated using formula (11):

[0150]

[0151] Among them, C lft is the intention cost of changing lane to the left, C kp The intention cost of keeping lane, C rt is the intention cost of changing lanes to the right, k is a coefficient, b l 、b k 、b r is a constant term;

[0152] The intention probability is calculated using formulas (12), (13) and (14):

[0153]

[0154]

[0155]

[0156] in, is the intention probability of changing lane to the left, To maintain the lane intention probability, is the intention probability of changing lane to the right.

[0157] In some embodiments of the present application, based on the actual vehicle data in different scenarios, the above mathematical model is subjected to Taylor series expansion or linear transformation, and the parameters of the mathematical model are solved using gradient descent or particle swarm algorithm, etc. The parameters refer to the above coefficients and constant terms.

[0158] The present application determines the historical intention results corresponding to each historical moment based on the driving state information of the target vehicle at multiple historical moments, and sets the influence of each historical intention result on the target prediction result through a forgetting factor, wherein the target prediction result is used to characterize the driving intention of the target vehicle. On the one hand, the above method does not need to be based on a deep learning model, and does not need to use high-definition image information, which reduces the requirements for computer computing power and image clarity, and saves the cost of predicting the driving intention of the target vehicle. On the other hand, the use of driving state data at multiple historical moments avoids the problems of inaccurate target prediction results and strong dependence of target prediction results on the accuracy of driving state data at that moment when predicting the driving trajectory based on driving state data at one moment, thereby improving the accuracy of target prediction results. In addition, the weight corresponding to the historical intention results corresponding to each historical moment is controlled by a forgetting factor to further improve the accuracy of the target prediction result. In addition, the mathematical model obtained based on the above method can be analyzed specifically for a specific scenario, so as to analyze and adjust the parameters of the mathematical model and achieve the purpose of optimizing the mathematical model. For example, according to the interactive scenario, the corresponding interactive parameters such as the longitudinal speed of the surrounding vehicle and the calculation formulas corresponding to the related parameters are added, such as formulas (4), (5) and (6), to optimize the mathematical model.

[0159] The present application also provides a vehicle driving intention prediction device, Figure 5 is a schematic diagram of the structure of a vehicle driving intention device provided in an embodiment of the present application, such as Figure 5 As shown, the vehicle driving intention prediction device 500 includes: a state acquisition module 501, a state determination module 502, a forgetting factor determination module 503 and an intention determination module 504. The state acquisition module 501 is used to obtain the driving state information of the target vehicle at multiple historical moments; the state determination module 502 is used to determine the historical intention result corresponding to each historical moment according to the driving state information of each historical moment in the multiple historical moments; the forgetting factor determination module 503 is used to determine the forgetting factor corresponding to each historical moment; the forgetting factor is used to characterize the influence degree of the historical intention result of the corresponding historical moment on the target prediction result; the target prediction result is used to characterize the driving intention of the target vehicle; the intention determination module 504 is used to determine the target prediction result according to the historical intention result and the forgetting factor corresponding to each historical moment in the multiple historical moments.

[0160] In some embodiments of the present application, the forgetting factor determination module 503 includes:

[0161] A first forgetting factor determination submodule is used to determine the time sequence of multiple historical moments;

[0162] The second forgetting factor determination submodule is used to determine the forgetting factor corresponding to each historical moment; wherein the influence degree represented by each forgetting factor increases in chronological order.

[0163] In some embodiments of the present application, the driving state information at each historical moment includes the lateral speed of the target vehicle, the heading angle of the target vehicle, the vehicle width of the target vehicle, the lane width of the current lane where the target vehicle is located, and the lateral position of the target vehicle relative to the current lane;

[0164] The intention determination module 504 includes:

[0165] A first intention determination submodule, for determining a lateral offset ratio based on a vehicle width, a lane width, and a lateral position corresponding to each historical moment;

[0166] The second intention determination submodule is used to determine the historical intention result corresponding to each historical moment based on the lateral offset ratio, the lateral speed and the heading angle.

[0167] In some embodiments of the present application, the above-mentioned device 500 further includes:

[0168] An interaction acquisition module is used to acquire the interaction information between the target vehicle and the surrounding vehicles of the target vehicle at the current moment; the surrounding vehicles include vehicles within a preset distance from the target vehicle;

[0169] An interaction determination module, used to determine a first auxiliary driving intention result according to the interaction information;

[0170] The interactive output module is used to output the driving intention of the target vehicle according to the first auxiliary driving intention result and the target prediction result.

[0171] In some embodiments of the present application, the above-mentioned interactive information includes at least one of momentum-related information of the surrounding vehicle, the longitudinal velocity of the surrounding vehicle, the longitudinal distance between the surrounding vehicle and the target vehicle, the target longitudinal velocity of the target vehicle and the target longitudinal acceleration of the target vehicle.

[0172] In some embodiments of the present application, the above-mentioned device 500 further includes:

[0173] A traffic acquisition module is used to acquire traffic information related to the target vehicle; the traffic information includes lane line type and / or traffic light information;

[0174] A traffic determination module, used to determine a second auxiliary driving intention result corresponding to the traffic information according to preset traffic rules;

[0175] The traffic output module is used to output the driving intention of the target vehicle according to the second auxiliary driving intention result and the target prediction result.

[0176] In some embodiments of the present application, the above-mentioned device 500 further includes:

[0177] A weather acquisition module, used to acquire current weather information of the area where the target vehicle is located; the current weather information includes at least one of rainfall and weather type;

[0178] A weather determination module, used to determine a third auxiliary driving intention result according to current weather information;

[0179] The weather output module is used to output the driving intention of the target vehicle according to the third auxiliary driving intention result and the target prediction result.

[0180] In some embodiments of the present application, the above-mentioned driving intention includes at least one of an intention to change lanes to the left, an intention to keep the lane, and an intention to change lanes to the right.

[0181] The device and method embodiments in the embodiments of this application are based on the same application concept.

[0182] Figure 6 The invention is a block diagram of an electronic device for predicting vehicle driving intention according to an exemplary embodiment.

[0183] The electronic device may be a server or a terminal device, and its internal structure diagram may be as follows: Figure 6 As shown. The electronic device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a vehicle driving intention prediction method is implemented.

[0184] Those skilled in the art will understand that Figure 6 The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present disclosure, and does not constitute a limitation on the electronic device to which the scheme of the present disclosure is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0185] An embodiment of the present application also provides an electronic device, comprising a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the above-mentioned vehicle driving intention prediction method.

[0186] An embodiment of the present application provides a vehicle, which includes the above-mentioned electronic device.

[0187] An embodiment of the present application also provides a computer storage medium, which can be set in a server to store at least one instruction, at least one program, code set or instruction set related to a vehicle driving intention prediction method in a method embodiment. The at least one instruction, the at least one program, the code set or instruction set is loaded and executed by the processor to implement the above-mentioned vehicle driving intention prediction method.

[0188] Optionally, in this embodiment, the storage medium may be located in at least one of the multiple network servers of the computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a mobile hard disk, a magnetic disk, or an optical disk.

[0189] In an exemplary embodiment, a computer program product is also provided, which includes a computer program stored in a readable storage medium. At least one processor of a computer device reads and executes the computer program from the readable storage medium, so that the computer device executes the vehicle driving intention prediction method of the embodiment of the present disclosure.

[0190] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0191] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0192] In the description provided herein, a large number of specific details are described. However, it is understood that the embodiments of the present application can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0193] Similarly, it should be understood that in order to streamline the present application and aid in understanding one or more of the various inventive aspects, in the above description of the exemplary embodiments of the present application, the various features of the present application are sometimes grouped together into a single embodiment, figure, or description thereof. However, this disclosed method should not be interpreted as reflecting the following intention: the claimed application requires more features than the features explicitly recited in each claim. More specifically, as reflected in the claims, the inventive aspects lie in less than all the features of the single embodiment disclosed above. Therefore, the claims that follow the specific embodiment are hereby expressly incorporated into the specific embodiment, with each claim itself serving as a separate embodiment of the present application.

[0194] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition they may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0195] In addition, those skilled in the art will appreciate that, although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the present application and form different embodiments. For example, in the claims, any one of the claimed embodiments can be used in any combination.

Claims

1. A vehicle driving intention prediction method, characterized in that: The method comprises: Acquire driving state information of the target vehicle at multiple historical moments; wherein the driving state information at each historical moment includes the lateral speed of the target vehicle, the heading angle of the target vehicle, the vehicle width of the target vehicle, the lane width of the current lane where the target vehicle is located, and the lateral position of the target vehicle relative to the current lane; Determining a lateral offset ratio based on the vehicle width, lane width, and lateral position corresponding to each of the historical moments; Determining a historical intention result corresponding to each of the historical moments based on the lateral offset ratio, the lateral speed and the heading angle; Determine the forgetting factor corresponding to each of the historical moments in chronological order; the influence degree represented by each of the forgetting factors increases in sequence in chronological order; the forgetting factor is used to represent the influence degree of the historical intention result of the corresponding historical moment on the target prediction result; the target prediction result is used to represent the driving intention of the target vehicle; The target prediction result is determined according to the historical intention result and the forgetting factor corresponding to each of the multiple historical moments.

2. The method according to claim 1, characterized in that The method further comprises: Acquire the interaction information between the target vehicle and surrounding vehicles of the target vehicle at the current moment; the surrounding vehicles include vehicles within a preset distance from the target vehicle; Determining a first auxiliary driving intention result according to the interactive information; The driving intention of the target vehicle is output according to the first auxiliary driving intention result and the target prediction result.

3. The method according to claim 2, characterized in that The interaction information includes at least one of momentum-related information of the surrounding vehicle, a longitudinal velocity of the surrounding vehicle, a longitudinal spacing between the surrounding vehicle and the target vehicle, a target longitudinal velocity of the target vehicle, and a target longitudinal acceleration of the target vehicle.

4. The method according to claim 1, characterized in that: The method further comprises: Acquire traffic information related to the target vehicle; the traffic information includes lane line type and / or traffic light information; Determining a second auxiliary driving intention result corresponding to the traffic information according to preset traffic rules; The driving intention of the target vehicle is output according to the second auxiliary driving intention result and the target prediction result.

5. The method according to claim 1, characterized in that The method further comprises: Acquire current weather information of the area where the target vehicle is located; the current weather information includes at least one of rainfall and weather type; determining a third auxiliary driving intention result according to the current weather information; The driving intention of the target vehicle is output according to the third auxiliary driving intention result and the target prediction result.

6. The method according to claim 1, characterized in that The driving intention includes at least one of an intention to change lanes to the left, an intention to keep the lane, and an intention to change lanes to the right.

7. A vehicle driving intention prediction device, characterized in that: The device comprises: A state acquisition module, used to acquire driving state information of a target vehicle at multiple historical moments; wherein the driving state information at each historical moment includes the lateral speed of the target vehicle, the heading angle of the target vehicle, the vehicle width of the target vehicle, the lane width of the current lane where the target vehicle is located, and the lateral position of the target vehicle relative to the current lane; a state determination module, configured to determine a lateral offset ratio based on a vehicle width, a lane width, and a lateral position corresponding to each of the historical moments; Determining a historical intention result corresponding to each of the historical moments based on the lateral offset ratio, the lateral speed and the heading angle; A forgetting factor determination module determines the forgetting factor corresponding to each of the historical moments in chronological order; the influence degree represented by each of the forgetting factors increases in chronological order; the forgetting factor is used to represent the influence degree of the historical intention result of the corresponding historical moment on the target prediction result; the target prediction result is used to represent the driving intention of the target vehicle; An intention determination module is used to determine the target prediction result based on the historical intention result corresponding to each of the multiple historical moments and the forgetting factor.

8. An electronic device, characterized in that: The device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the vehicle driving intention prediction method as described in any one of claims 1-6.

9. A vehicle, characterized in that: The vehicle comprises the electronic device according to claim 8.

Images