CN110908375A - Method and device for acquiring lane change decision information, storage medium and vehicle - Google Patents

Method and device for acquiring lane change decision information, storage medium and vehicle Download PDF

Info

Publication number
CN110908375A
CN110908375A CN201911114524.2A CN201911114524A CN110908375A CN 110908375 A CN110908375 A CN 110908375A CN 201911114524 A CN201911114524 A CN 201911114524A CN 110908375 A CN110908375 A CN 110908375A
Authority
CN
China
Prior art keywords
lane change
vehicle
time
state information
prediction
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
CN201911114524.2A
Other languages
Chinese (zh)
Inventor
周奕达
丁曙光
任冬淳
连世奇
付圣
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Sankuai Online Technology Co Ltd
Original Assignee
Beijing Sankuai Online Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Sankuai Online Technology Co Ltd filed Critical Beijing Sankuai Online Technology Co Ltd
Priority to CN201911114524.2A priority Critical patent/CN110908375A/en
Publication of CN110908375A publication Critical patent/CN110908375A/en
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/0088—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots characterized by the autonomous decision making process, e.g. artificial intelligence, predefined behaviours
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/02—Control of position or course in two dimensions
    • G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
    • G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
    • G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G1/00—Traffic control systems for road vehicles
    • G08G1/01—Detecting movement of traffic to be counted or controlled
    • G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
    • G08G1/0125—Traffic data processing
    • G08G1/0133—Traffic data processing for classifying traffic situation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Health & Medical Sciences (AREA)
  • Business, Economics & Management (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Game Theory and Decision Science (AREA)
  • Medical Informatics (AREA)
  • Chemical & Material Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Traffic Control Systems (AREA)

Abstract

The present disclosure relates to a method, an apparatus, a storage medium, and a vehicle for obtaining lane change decision information, including: acquiring running state information of the vehicle at the current moment; determining lane change behaviors of the vehicle at the current time according to the running state information at the current time, and predicting the running state information of the vehicle at a first prediction time after the lane change behaviors are carried out on the vehicle aiming at each lane change behavior at the current time; determining lane change behaviors of the vehicle at the first prediction time according to the running state information at the first prediction time, and predicting running state information of the vehicle at a second prediction time after the lane change behaviors are performed for each lane change behavior at the first prediction time; and determining lane change decision information of the vehicle according to the driving state information at the current time, the lane change behaviors which can be performed by the vehicle at the current time, the driving state information of the vehicle at each prediction time and the lane change behaviors which can be performed by the vehicle at each prediction time.

Description

Method and device for acquiring lane change decision information, storage medium and vehicle
Technical Field
The present disclosure relates to the field of unmanned driving, and in particular, to a method, an apparatus, a storage medium, and a vehicle for acquiring lane change decision information.
Background
The behavior decision is a very important link in unmanned driving, and the link determines the behaviors of lane changing, parking, acceleration, deceleration and the like of the unmanned vehicle. The lane change decision provides accurate reference control for the lane on which the vehicle runs.
In the prior art, lane change decision is determined based on a rule, a lane change decision is determined based on a model and the like, however, in the lane change decision determining mode based on the rule, the driving environment of a vehicle is difficult to be comprehensively set according to the rule, so that the flexibility and the accuracy are poor; when determining lane change decision based on the model, the model is trained through a large amount of real driving data, and the applicability is not high.
Disclosure of Invention
The invention aims to provide a method, a device, a storage medium and a vehicle for acquiring lane change decision information, which are accurate, flexible and high in applicability.
In order to achieve the above object, according to a first aspect of the present disclosure, there is provided a method of acquiring lane change decision information, the method including:
acquiring running state information of the vehicle at the current moment, wherein the running state information comprises the vehicle state information and the surrounding environment information of the vehicle;
determining lane change behaviors which can be performed by the vehicle at the current time according to the running state information at the current time, and predicting the running state information at a first prediction time after the lane change behaviors are performed by the vehicle for each lane change behavior which can be performed by the vehicle at the current time;
determining lane change behaviors that the host vehicle can perform at the first prediction time according to the driving state information at the first prediction time, and predicting the driving state information at a second prediction time after the host vehicle performs the lane change behaviors for each lane change behavior that the host vehicle can perform at the first prediction time;
and determining lane change decision information of the vehicle according to the running state information at the current time, the lane change behaviors which can be carried out by the vehicle at the current time, the running state information of the vehicle at each prediction time and the lane change behaviors which can be carried out by the vehicle at each prediction time.
Optionally, the number of the predicted time is N, where N is a positive integer greater than 2, the first predicted time is initially the closest predicted time to the current time among the N predicted times, and the second predicted time is the closest predicted time to the first predicted time among the N predicted times;
after predicting the travel state information of the host vehicle at the second prediction time, the method further includes:
and a step of, when there is a predicted time after the second predicted time, setting the second predicted time as a new first predicted time, and executing the step of determining a lane change behavior that can be performed by the host vehicle at the first predicted time again based on the travel state information at the first predicted time, and predicting the travel state information at the second predicted time after the lane change behavior of the host vehicle is performed for each lane change behavior that can be performed by the host vehicle at the first predicted time until the travel state information at each predicted time is predicted.
Optionally, the determining lane change decision information of the host vehicle according to the driving state information at the current time, the lane change behavior that can be performed by the host vehicle at the current time, the driving state information of the host vehicle at each prediction time, and the lane change behavior that can be performed by the host vehicle at each prediction time includes:
obtaining a target tree according to the running state information at the current moment, the lane change behavior which can be performed by the vehicle at the current moment, the running state information of the vehicle at each prediction moment and the lane change behavior which can be performed by the vehicle at each prediction moment, wherein a root node of the target tree corresponds to the current moment and represents the running state information of the vehicle at the current moment, the target tree comprises M layers of sub-nodes, each layer of sub-node corresponds to one prediction moment, and M is the total number of the prediction moments; each child node with the same father node corresponds to each lane change behavior which can be performed by the vehicle at the time corresponding to the layer where the father node is located, and represents the driving state information of the vehicle at the time corresponding to the layer where the child node is located after the vehicle performs the corresponding lane change behavior;
determining a target parameter of each node path in the target tree, wherein the size of the target parameter is used for distinguishing the advantages and disadvantages among the node paths;
determining an optimal node path according to the target parameters;
and determining lane change decision information of the vehicle according to the optimal node path.
Optionally, the determining a target parameter of each node path in the target tree includes:
determining parameters for each node in the target tree from one or more of the following items of information: lane change behavior information, lane change state information, result information after predicting lane change behavior, and continuous information between the lane change behavior of the current node and the lane change behavior of a father node of the current node;
for each node path, determining a weighted sum of the parameters of each node in the node path as a target parameter of the node path.
Optionally, the determining, according to the driving state information at the current time, a lane change behavior that the host vehicle can perform at the current time includes:
predicting the motion trail of surrounding obstacles through a trail prediction model according to the surrounding environment information of the vehicle at the current moment, and determining lane change behaviors which can be performed by the vehicle at the current moment through a lane change behavior determination model according to the motion trail and the state information of the vehicle at the current moment;
the determining, according to the driving state information at the first prediction time, a lane change behavior that can be performed by the host vehicle at the first prediction time includes:
and predicting the motion trail of the surrounding obstacle through the trail prediction model according to the surrounding environment information of the vehicle at the first prediction time, and determining lane change behaviors which can be performed by the vehicle at the first prediction time through the lane change behavior determination model according to the motion trail and the vehicle state information at the first prediction time.
Optionally, the method further comprises:
determining road condition information of a road section where the vehicle is located at the current moment;
if the vehicle is in a congested road section at the current moment, the track prediction model is a first track prediction model, and the lane change behavior determination model is a first lane change behavior determination model;
and if the vehicle is located at the highway section at the current moment, the track prediction model is a second track prediction model, and the lane change behavior determination model is a second lane change behavior determination model, wherein the second track prediction model is different from the first track prediction model, and the second lane change behavior determination model is different from the first lane change behavior determination model.
Optionally, the first trajectory prediction model is implemented based on a mixed density network, and the first lane change behavior determination model is implemented based on a convolutional neural network; the second trajectory prediction model is realized based on a long-term and short-term memory network, and the second lane change behavior determination model is realized based on a Bayesian network.
According to a second aspect of the present disclosure, there is provided an apparatus for acquiring lane change decision information, the apparatus comprising:
the device comprises a first acquisition module, a second acquisition module and a control module, wherein the first acquisition module is used for acquiring the driving state information of the vehicle at the current moment, and the driving state information comprises the vehicle state information and the surrounding environment information of the vehicle;
a first prediction module, configured to determine, according to the driving state information at the current time, a lane change behavior that can be performed by the host vehicle at the current time, and predict, for each lane change behavior that can be performed by the host vehicle at the current time, the driving state information at a first prediction time after the lane change behavior is performed by the host vehicle;
a second prediction module configured to determine, based on the travel state information at the first prediction time, lane change behaviors that can be performed by the host vehicle at the first prediction time, and predict, for each lane change behavior that can be performed by the host vehicle at the first prediction time, the travel state information at a second prediction time after the lane change behavior of the host vehicle has been performed;
the first determining module is used for determining lane change decision information of the vehicle according to the running state information at the current time, the lane change behaviors which can be carried out by the vehicle at the current time, the running state information of the vehicle at each prediction time and the lane change behaviors which can be carried out by the vehicle at each prediction time.
Optionally, the number of the predicted time is N, where N is a positive integer greater than 2, the first predicted time is initially the closest predicted time to the current time among the N predicted times, and the second predicted time is the closest predicted time to the first predicted time among the N predicted times;
the device further comprises:
and a triggering module configured to, after the second prediction module predicts the travel state information of the host vehicle at a second prediction time, if a prediction time is left after the second prediction time, set the second prediction time as a new first prediction time, and trigger the second prediction model to determine a lane change behavior that the host vehicle can perform at the first prediction time, based on the travel state information at the first prediction time, and predict the travel state information at the second prediction time after the lane change behavior of the host vehicle is performed for each lane change behavior that the host vehicle can perform at the first prediction time, until the travel state information of the host vehicle at each prediction time is predicted.
Optionally, the first determining module includes:
a first processing sub-module, configured to obtain a target tree according to the driving state information at the current time, a lane change behavior that can be performed by a host vehicle at the current time, the driving state information at each prediction time, and a lane change behavior that can be performed by a host vehicle at each prediction time, where a root node of the target tree corresponds to the current time and represents the driving state information of the host vehicle at the current time, the target tree includes M layers of sub-nodes, each layer of sub-node corresponds to a prediction time, where M is a total number of prediction times; each child node with the same father node corresponds to each lane change behavior which can be performed by the vehicle at the time corresponding to the layer where the father node is located, and represents the driving state information of the vehicle at the time corresponding to the layer where the child node is located after the vehicle performs the corresponding lane change behavior;
the first determining submodule is used for determining a target parameter of each node path in the target tree, and the size of the target parameter is used for distinguishing the advantages and disadvantages among the node paths;
the second determining submodule is used for determining an optimal node path according to the target parameter;
and the third determining submodule is used for determining lane change decision information of the vehicle according to the optimal node path.
Optionally, the first determining sub-module includes:
a fourth determining submodule, configured to determine a parameter of each node in the target tree according to one or more of the following information items: lane change behavior information, lane change state information, result information after predicting lane change behavior, and continuous information between the lane change behavior of the current node and the lane change behavior of a father node of the current node;
and the fifth determining submodule is used for determining the weighted sum of the parameters of each node in each node path as the target parameter of the node path.
Optionally, the first prediction module comprises:
a sixth determining submodule, configured to predict a motion trajectory of a peripheral obstacle through a trajectory prediction model according to the information about the environment around the host vehicle at the current time, and determine a lane change behavior that the host vehicle can perform at the current time through a lane change behavior determination model according to the motion trajectory and the state information of the host vehicle at the current time;
the second prediction module comprises:
and a seventh determining submodule, configured to predict, according to the information about the environment around the host vehicle at the first prediction time, a motion trajectory of a peripheral obstacle through the trajectory prediction model, and determine, according to the motion trajectory and the host vehicle state information at the first prediction time, a lane change behavior that the host vehicle can perform at the first prediction time through the lane change behavior determination model.
Optionally, the apparatus further comprises:
the second determining module is used for determining the road condition information of the road section where the vehicle is located at the current moment; if the vehicle is in a congested road section at the current moment, the track prediction model is a first track prediction model, and the lane change behavior determination model is a first lane change behavior determination model; and if the vehicle is located at the highway section at the current moment, the track prediction model is a second track prediction model, and the lane change behavior determination model is a second lane change behavior determination model, wherein the second track prediction model is different from the first track prediction model, and the second lane change behavior determination model is different from the first lane change behavior determination model.
Optionally, the first trajectory prediction model is implemented based on a mixed density network, and the first lane change behavior determination model is implemented based on a convolutional neural network; the second trajectory prediction model is realized based on a long-term and short-term memory network, and the second lane change behavior determination model is realized based on a Bayesian network.
According to a third aspect of the present disclosure, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of any of the methods of the first aspect described above.
According to a fourth aspect of the present disclosure, there is provided a vehicle comprising:
a memory having a computer program stored thereon;
a processor for executing the computer program in the memory to implement the steps of the method of any of the first aspects above.
In the above-described aspect, the lane change behavior and the travel state information at the plurality of prediction times may be determined based on the vehicle state information of the vehicle at the current time and the vehicle surrounding environment information, so that the lane change decision information of the vehicle may be determined based on the travel state information at the current time, the lane change behavior that the vehicle can perform at the current time, the travel state information of the vehicle at each prediction time, and the lane change behavior that the vehicle can perform at each prediction time. According to the technical scheme, on one hand, when the lane change behavior which can be carried out by the vehicle at each prediction time and the running state information at each prediction time are determined, the prediction is carried out based on the running state information at the previous prediction time, so that the continuity and the integrity of the corresponding running state information at a plurality of prediction times can be ensured, the accuracy of lane change decision information is improved, and the flexibility of obtaining the lane change decision information can be effectively improved. On the other hand, the lane change decision information is determined based on the vehicle state information and the vehicle surrounding environment information, so that the current driving environment can be adapted when the lane change decision information is acquired, and the adaptability of acquiring the lane change decision information is improved.
Additional features and advantages of the disclosure will be set forth in the detailed description which follows.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description serve to explain the disclosure without limiting the disclosure. In the drawings:
fig. 1 is a flowchart of a method of obtaining lane change decision information provided according to an embodiment of the present disclosure;
FIG. 2 is a flow diagram of an exemplary embodiment of determining lane change decision information for a host vehicle based on the travel state information at a current time, the lane change behavior that the host vehicle may perform at the current time, the travel state information for the host vehicle at each predicted time, and the lane change behavior that the host vehicle may perform at each predicted time;
FIG. 3 is a schematic diagram of a target tree provided in accordance with one embodiment of the present disclosure;
FIG. 4 is a schematic illustration of a vehicle and a roadway provided in accordance with an embodiment of the present disclosure;
FIG. 5 is a block diagram of an apparatus for obtaining lane change decision information according to an embodiment of the present disclosure;
FIG. 6 is a block diagram of a vehicle shown in accordance with an exemplary embodiment;
FIG. 7 is a block diagram of a vehicle shown in accordance with an exemplary embodiment.
Detailed Description
The following detailed description of specific embodiments of the present disclosure is provided in connection with the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the present disclosure, are given by way of illustration and explanation only, not limitation.
Fig. 1 is a flowchart illustrating a method for obtaining lane change decision information according to an embodiment of the present disclosure. As shown in fig. 1, the method may include:
at S11, the driving state information of the host vehicle at the current time is acquired, and the driving state information includes the host vehicle state information and the host vehicle surrounding environment information.
The own vehicle state information may include information such as a position, a heading, a speed, an acceleration and the like of the own vehicle, and the own vehicle surrounding environment information may include one or more of surrounding vehicle information, surrounding pedestrian information and surrounding non-motor vehicle information. The vehicle state information may be obtained by a sensor mounted on the vehicle, the ambient information of the vehicle may be obtained by a camera mounted on the vehicle, and thus obtained by performing image processing and analysis on the image, or obtained by message interconnection of the internet of vehicles, which is not limited in this disclosure.
At S12, the lane change behavior that the host vehicle can perform at the current time is specified based on the driving state information at the current time, and the driving state information at the first prediction time after the lane change behavior of the host vehicle is predicted for each lane change behavior that the host vehicle can perform at the current time.
As an example, determining the lane change behaviors that the host vehicle can perform at the current time may be determining the feasibility of each of the lane change behaviors that the host vehicle can perform at the current time, e.g., determining the feasibility of performing a left lane change of the host vehicle at the current time to be 40%, the feasibility of performing a right lane change to be 20%, and the feasibility of keeping the current lane to be 40%.
Alternatively, the driving state of the host vehicle may be modeled based on a Markov Decision Process (MDP) to obtain an MDP state transition equation corresponding to the host vehicle, so that the state information of the host vehicle at the first prediction time after the lane change behavior of the host vehicle is performed may be derived based on the MDP state transition equation. In an example, the vehicle surrounding environment information at the first prediction time may be predicted based on the acquired vehicle surrounding environment information.
At S13, the lane change behavior that the host vehicle can perform at the first prediction time is determined based on the travel state information at the first prediction time, and the travel state information at the second prediction time after the lane change behavior of the host vehicle has been performed is predicted for each lane change behavior that the host vehicle can perform at the first prediction time.
However, as is clear from S12 above, if the travel state information at the first prediction time corresponds one-to-one to the lane change behavior that the host vehicle can perform at the current time, in this step, the travel state information corresponding to each lane change behavior at the first prediction time (that is, the travel state information at the first prediction time after the predicted host vehicle performed the lane change behavior) may be used to determine the lane change behavior that the host vehicle can perform at the first prediction time from the travel state information corresponding to the lane change behavior, and the travel state information at the second prediction time after the lane change behavior is performed for each lane change behavior that the host vehicle can perform at the first prediction time may be predicted.
In the above example, the lane change behavior that the host vehicle can perform at the current time includes left lane change, right lane change, and current lane keeping, and the driving state information at the first prediction time includes driving state information at the first prediction time after the host vehicle performs the left lane change, driving state information at the first prediction time after the host vehicle performs the right lane change, and driving state information at the first prediction time after the host vehicle keeps the current lane, so that the driving state information at the second prediction time can be determined based on the driving state information at the first prediction time. The manner of predicting the driving state information at the second prediction time after the lane change behavior of the vehicle is performed is the same as that described above, and is not described herein again.
At S14, lane change decision information of the host vehicle is determined based on the travel state information at the current time, the lane change behavior that can be performed by the host vehicle at the current time, the travel state information of the host vehicle at each predicted time, and the lane change behavior that can be performed by the host vehicle at each predicted time.
In the above-described aspect, the lane change behavior and the travel state information at the plurality of prediction times may be determined based on the vehicle state information of the vehicle at the current time and the vehicle surrounding environment information, so that the lane change decision information of the vehicle may be determined based on the travel state information at the current time, the lane change behavior that the vehicle can perform at the current time, the travel state information of the vehicle at each prediction time, and the lane change behavior that the vehicle can perform at each prediction time. According to the technical scheme, on one hand, when the lane change behavior which can be carried out by the vehicle at each prediction time and the running state information at each prediction time are determined, the prediction is carried out based on the running state information at the previous prediction time, so that the continuity and the integrity of the corresponding running state information at a plurality of prediction times can be ensured, the accuracy of lane change decision information is improved, and the flexibility of obtaining the lane change decision information can be effectively improved. On the other hand, the lane change decision information is determined based on the vehicle state information and the vehicle surrounding environment information, so that the current driving environment can be adapted when the lane change decision information is acquired, and the adaptability of acquiring the lane change decision information is improved.
Optionally, the predicted time is N, where N is a positive integer greater than 2, the first predicted time is initially the predicted time closest to the current time among the N predicted times, and the second predicted time is the predicted time closest to the first predicted time among the N predicted times. The number of the predicted time can be determined according to the predicted total time and the predicted step length, and for example, if the predicted total time is 4s and the predicted step length is 1s, it is determined that the predicted time is 4. Wherein, if the current time is 0s, the first predicted time is 1s, and the second predicted time is 2 s.
After predicting the travel state information of the host vehicle at the second prediction time, the method further includes:
and a step of, when there is a predicted time after the second predicted time, setting the second predicted time as a new first predicted time, and executing the step of determining a lane change behavior that can be performed by the host vehicle at the first predicted time again based on the travel state information at the first predicted time, and predicting the travel state information at the second predicted time after the lane change behavior of the host vehicle is performed for each lane change behavior that can be performed by the host vehicle at the first predicted time until the travel state information at each predicted time is predicted.
As in the above example, if the determined prediction times are 4, then after the driving state information of the host vehicle at the second prediction time (i.e., the 2 nd time) is predicted, and then the prediction times (i.e., the 3 rd and the 4 th time) are left after the second prediction time, the second prediction time may be set as a new first prediction time, that is, the lane change behavior that the host vehicle can perform at the 2 nd time may be determined based on the driving state information at the new first prediction time (i.e., the 2 nd time), and the driving state information at the new second prediction time (i.e., the 3 rd time) may be predicted after the lane change behavior is performed by the host vehicle for each lane change behavior that the host vehicle can perform at the 2 nd time.
The running state information at the 4 th time can be predicted in turn by the loop mode, so that the running state information of the vehicle at each predicted time can be predicted.
In the technical scheme, by predicting a plurality of predicted times after the current time, the running state information of the vehicle at the plurality of predicted times can be predicted, so that the states of the vehicle at the plurality of predicted times after the current time can be predicted, the lane change decision information of the vehicle can be determined, the state of the vehicle in a future time period can be considered more comprehensively, and more accurate and comprehensive data reference is provided for determining accurate lane change decision information.
Alternatively, an exemplary embodiment of the determining lane change decision information of the host vehicle according to the driving state information at the current time, the lane change behavior that can be performed by the host vehicle at the current time, the driving state information of the host vehicle at each predicted time, and the lane change behavior that can be performed by the host vehicle at each predicted time is as follows, as shown in fig. 2, and the step may include:
at S21, obtaining a target tree according to the driving state information at the current time, the lane change behavior that can be performed by the vehicle at the current time, the driving state information at each predicted time, and the lane change behavior that can be performed by the vehicle at each predicted time, wherein a root node of the target tree corresponds to the current time and represents the driving state information at the current time, the target tree includes M layers of sub-nodes, each layer of sub-node corresponds to a predicted time, and M is the total number of predicted times; each child node having the same parent node corresponds to each lane change behavior that the host vehicle can perform at the time corresponding to the layer where the parent node is located, and represents the driving state information at the time corresponding to the layer where the child node is located after the host vehicle performs the corresponding lane change behavior.
The process of constructing the target tree is described in detail below based on a specific embodiment.
For example, as shown in fig. 3, the current time is 0s, and the corresponding node is b0 as the root node of the target tree. Then, the child nodes of b0 may be determined according to each lane change behavior that the host vehicle can perform, which is determined at the current time, wherein each lane change behavior corresponds to a child node. For example, it is determined that the lane change behaviors that the host vehicle can perform at the current time are respectively a left lane change, a current lane keeping state, and a right lane change, then b0 corresponds to 3 child nodes, as shown in b1, b2, and b3 in fig. 3, and the time corresponding to the layer where b1, b2, and b3 are located is the first prediction time t1, that is, the 1 st s. The b1 may correspond to the left lane change behavior that the host vehicle can perform at the time t0 (i.e., the 0 th s at the current time) corresponding to the layer on which the b0 is located, and the child node b1 is used to indicate the driving state information of the host vehicle at the first predicted time t1 (i.e., the 1 st s) after performing the left lane change behavior. By analogy, child nodes b2 and b3 can be constructed, respectively, as shown in fig. 3.
Thereafter, a child node corresponding to the second prediction time t2 (i.e., 2s) may be created according to the child node of the layer of the first prediction time t 1. Wherein, the child nodes of the t1 level can be sequentially selected and used as new root nodes, and the operation of creating child nodes is executed. After traversing the child nodes at the level t1, the child nodes corresponding to the second prediction time t2 are continuously traversed according to the above manner, so that the target tree is constructed according to the driving state information at each prediction time.
In S22, a target parameter of each node path in the target tree is determined, and the target parameter is used for distinguishing between the node paths. The corresponding relation between the size of the target parameter and the quality of the node path can be set according to the specific determination mode of the target parameter.
In S23, an optimal node path is determined based on the target parameters.
In S24, lane change decision information of the host vehicle is determined from the optimal node path.
Each node path comprises a plurality of nodes, and each node corresponds to a prediction time, so that after the optimal node path is determined, the lane change behavior corresponding to each sub-node in the optimal node path can be used as the lane change decision information of the vehicle to control the vehicle to run.
For example, as shown in fig. 3, if the determined optimal node path is b0-b1-b11-b111-b1111, the corresponding lane change decision information may be "left lane change-left lane change", and thus, the driving of the host vehicle may be controlled according to the lane change decision information.
Optionally, in the process of controlling the vehicle to run according to the determined lane change decision information of the vehicle, if new lane change decision information is determined, the vehicle is controlled to run based on the new lane change decision information, so as to ensure real-time performance and accuracy of vehicle running control.
In the above-described aspect, the target tree is obtained based on the travel state information at the current time, the lane change behavior that can be performed by the host vehicle at the current time, the travel state information at each predicted time, and the lane change behavior that can be performed by the host vehicle at each predicted time, so that the travel state information of the host vehicle in the future period can be comprehensively represented based on the target tree. Therefore, each node path can be intuitively and simply determined based on the target tree, so that the optimal node path is determined by comparing the advantages and disadvantages among the node paths, the lane change decision information of the vehicle is further determined, the determination efficiency of the lane change decision information can be improved, and the real-time performance of vehicle running control is improved.
Optionally, the determining a target parameter of each node path in the target tree includes:
determining parameters for each node in the target tree from one or more of the following items of information: lane change behavior information, lane change state information, result information after predicting lane change behavior, and continuous information between the lane change behavior of the current node and the lane change behavior of the father node of the current node.
The lane change behavior information is used for representing lane change behaviors corresponding to the current node, such as left lane change, right lane change and the like, parameter values corresponding to different lane change behavior information can be set according to actual use scenes, and the parameter values corresponding to different lane change behavior information can be the same or different; the parameter values corresponding to the same lane change behavior information in different application scenes may be the same or different, for example, in a scene in which the vehicle runs to the left and a scene in which the vehicle runs to the right, the parameter values corresponding to the left lane change may be the same or different, and this disclosure does not limit this.
The lane change state information may be used to indicate the distance between the position of the host vehicle under the node and the target position of the corresponding lane change behavior. When determining the lane change behavior that can be performed by the vehicle, a target position corresponding to the lane change behavior may be determined, and the target position may be a position of the vehicle corresponding to the successful lane change behavior. If the lane change action can be carried out, the target position is in the current lane; if the lane change action which can be carried out is left lane change, the target position is positioned in a left lane of the current lane; if the lane change action is right lane change, the target position is located in the right lane of the current lane. Unless otherwise specified, "left" and "right" in the present disclosure are both determined based on the current traveling direction of the host vehicle.
For example, as shown in fig. 4, the driving state information at the time b1, where S represents the host vehicle and the L point position represents the target position of the left lane change, when determining the parameter value of the lane change state information corresponding to b1, the distance determination from the target position L point may be determined according to the current position of the host vehicle S. As an example, the parameter value of the lane change state information may be an inverse of the distance, and the parameter value of the lane change state information is smaller the farther the distance of the host vehicle position from the target position is. As another example, the parameter value of the lane change state information may be determined according to a distance range in which the distance is located, wherein a correspondence relationship between the distance range and the parameter value of the lane change state information may be set in advance, and when the distance between the current position of the host vehicle S and the target position L is determined, the parameter value of the lane change state information may be directly calculated based on the correspondence relationship.
The result information after predicting the lane change behavior is used for indicating whether the lane change behavior is successful or not, and indicating that the lane change behavior is successful when the vehicle changes lanes, the parameter value of the result information may be 1, and when the lane change of the vehicle is not completed, if the included angle between the current heading of the vehicle and the lane direction is too large, the vehicle may be in the process of changing lanes, and at this time, the parameter value of the result information may be 0.
When determining the parameter value of the continuous information between the lane change behavior of the current node and the lane change behavior of the parent node of the current node, taking b11 in fig. 3 as an example, the lane change behavior corresponding to b11 is left lane change, and the lane change behavior corresponding to the parent node b1 of b11 is left lane change, that is, the lane change behavior representing the current node is continuous with the lane change behavior of the parent node of the current node, and then the parameter value of the continuous information may be 1. If the lane change behavior corresponding to b12 is terminated, and the lane change behavior corresponding to the parent node b1 of b12 is left lane change, that is, the lane change behavior of the current node is not continuous with the lane change behavior of the parent node of the current node, the parameter value of the continuous information may be 0. Wherein, the parameter value of the continuous information between the lane change behaviors of the parent-child nodes can be preset.
When determining the parameter value of the continuous information corresponding to b1, the parameter value may be set to a default value, or may be determined based on the continuity between the lane change behavior corresponding to b1 and the lane change behavior represented by the last lane change decision information.
The values of the above parameter values are merely exemplary, and do not limit the present disclosure. When determining a parameter of a node in the target tree from one of the above information items, a parameter value of the information item may be directly determined as a parameter of the node. When determining parameters of a node in the target tree according to multiple information items, the weighted sum of parameter values of multiple information items corresponding to the node can be determined as the parameters of the node, for example, when determining the parameters of the node according to lane change behavior information and lane change state information, the weighted sum can be performed based on the weighted values corresponding to the lane change behavior information and the lane change state information, and the parameter values of the lane change behavior information and the lane change state information, and the result of the weighted sum is determined as the parameters of the node, so that the comprehensiveness and accuracy of the parameter determination of the node can be ensured, and accurate data support can be provided for determining lane change decision information.
For each node path, determining a weighted sum of the parameters of each node in the node path as a target parameter of the node path.
Alternatively, a weight value corresponding to each layer of sub-nodes in the target tree may be preset, for example, as shown in fig. 3, for convenience of illustration, a complete node path of only one sub-branch of b1 is shown in fig. 3, other node paths are not shown, a first layer of sub-nodes corresponds to the first prediction time t1, and a corresponding weight value is α 1, for example, α 1, α 2, α 3, and α 4 are respectively used to represent weights corresponding to the layer of sub-nodes where t1, t2, t3, and t4 are located, when a target parameter of a node path is determined, parameters of each node on the node path are weighted and summed, as shown in fig. 3, the node b0-b1-b11-b111-b1111 is formed as a node path, and a target parameter corresponding to the node path is:
Q＝q0+q1*α1+q11*α2+q111*α3+q1111*α4。
q0 represents a parameter corresponding to b0, q1 represents a parameter corresponding to b1, q11 represents a parameter corresponding to b11, q111 represents a parameter corresponding to b111, and q1111 represents a parameter corresponding to b 1111.
Optionally, when the weight value corresponding to each layer of child nodes in the target tree is set, the weight value corresponding to the leaf node from the root node in the target tree is reduced, that is, α 1> α 2> α 3> α 4.
In the technical scheme, the target parameters of the node path can be determined according to the parameters of each node contained in the node path by determining the parameters of each node in the target tree, so that the accuracy of the target parameters of the node path can be ensured, the advantages and the disadvantages of the node path can be accurately characterized, accurate data support is provided for subsequent determination of the optimal node path and lane change decision information based on the target parameters, and the accuracy of the lane change decision information is improved.
Optionally, the determining, according to the driving state information at the current time, a lane change behavior that the host vehicle can perform at the current time includes:
predicting the motion trail of surrounding obstacles through a trail prediction model according to the surrounding environment information of the vehicle at the current moment, and determining lane change behaviors which can be performed by the vehicle at the current moment through a lane change behavior determination model according to the motion trail and the state information of the vehicle at the current moment;
the determining, according to the driving state information at the first prediction time, a lane change behavior that can be performed by the host vehicle at the first prediction time includes:
and predicting the motion trail of the surrounding obstacle through the trail prediction model according to the surrounding environment information of the vehicle at the first prediction time, and determining lane change behaviors which can be performed by the vehicle at the first prediction time through the lane change behavior determination model according to the motion trail and the vehicle state information at the first prediction time.
The surrounding obstacle may be an object that affects the unmanned driving of the host vehicle, such as a pedestrian, a vehicle, a non-motor vehicle, or the like around the host vehicle. In this embodiment, the motion trajectory of the surrounding obstacle can be dynamically predicted by the trajectory prediction model. And when the lane change behavior which can be carried out by the vehicle at the current moment is determined, the motion trail of the surrounding obstacles is considered in display, so that the influence of the motion trail of the surrounding obstacles on the driving of the vehicle is effectively reduced, and the reliability and the stability of the lane change decision information are improved.
Optionally, the method further comprises:
and determining road condition information of the road section where the vehicle is located at the current moment, wherein the road condition information of the road section where the vehicle is located at the current moment can be determined by combining map information and/or road network information. For example, the road condition information is obtained by determining a road segment where the vehicle is located by map information and/or road network information according to the position information of the vehicle at the current time.
If the vehicle is in a congested road section at the current moment, the track prediction model is a first track prediction model, and the lane change behavior determination model is a first lane change behavior determination model;
and if the vehicle is located at the highway section at the current moment, the track prediction model is a second track prediction model, and the lane change behavior determination model is a second lane change behavior determination model, wherein the second track prediction model is different from the first track prediction model, and the second lane change behavior determination model is different from the first lane change behavior determination model.
In the above embodiment, by determining the road condition information of the road segment where the vehicle is located at the current time, when determining the lane change behavior that the vehicle can perform at the current time based on the driving state information of the vehicle at the current time, the trajectory prediction model and the lane change behavior determination model that are matched with the road condition information of the road segment at the current time may be selected, so that different predictions may be performed according to different scenes when obtaining the lane change decision information, which may improve the accuracy of the lane change decision information and widen the applicability of the method for obtaining the lane change decision information to different scenes.
Optionally, the first trajectory prediction model is implemented based on a Mixed Density Network (MDN), and the first lane change behavior determination model is implemented based on a Convolutional Neural Network (CNN).
In the congested road section, behaviors of vehicles, pedestrians and non-motor vehicles in the surrounding environment are complex and changeable, and the complex and changeable behaviors have very important influence on determining lane change decision information. Therefore, in this case, an MDN model is employed as a trajectory prediction model to predict the movement trajectory of the surrounding obstacle. The method comprises the steps of inputting the surrounding environment information of the vehicle at the current moment into an MDN model, and predicting the probability distribution thermodynamic diagrams of the positions of all obstacles in the surrounding environment information of the vehicle, so that various possible motion trails of the surrounding obstacles can be predicted, and comprehensive prediction information can be provided for a lane change behavior determination model.
And the lane change behavior determination model under the scene is realized based on a convolutional neural network, and the convolutional neural network model can be trained in a data set training mode, so that the feasibility of each possible lane change behavior can be output after convolution is performed based on the probability distribution thermodynamic diagram output by the MDN model. The lane change behavior is determined through the convolutional neural network, driving behavior data of a driver can be used as training data, lane change decision information can be flexibly processed based on a mode of simulating driver driving, and therefore the problem that a rule-based lane change decision method in the prior art is too conservative and is difficult to adapt to congested road sections is solved.
The second track prediction model is realized based on a Long Short-Term Memory network (LSTM) model, and the second lane change behavior determination model is realized based on a Bayesian Network (BN).
The LSTM model can predict the movement track with the maximum possibility of the obstacle, and each node on the network has the corresponding physical meaning in the BN model, so that the reasoning process based on the BN model can be completely explained.
When the vehicle is in a high-speed road section, unmanned driving needs to ensure the stability and safety of vehicle lane changing. Since the jitter of the lane change decision information can cause the vehicle to jitter under the highway section, danger is caused. Therefore, in the scene, when determining the lane change behavior that the vehicle can perform at the current time, the lane change behavior of the vehicle is predicted based on the motion trail with the highest possibility of the surrounding obstacles and the vehicle state information, so that the stability and the safety of the obtained lane change decision information are ensured, and the safety of unmanned driving is further ensured.
The present disclosure also provides an apparatus for acquiring lane change decision information, as shown in fig. 5, where the apparatus 10 includes:
a first obtaining module 100, configured to obtain driving state information of a host vehicle at a current time, where the driving state information includes host vehicle state information and host vehicle surrounding environment information;
a first prediction module 200, configured to determine, according to the driving state information at the current time, a lane change behavior that can be performed by the host vehicle at the current time, and predict, for each lane change behavior that can be performed by the host vehicle at the current time, the driving state information at a first prediction time after the lane change behavior is performed by the host vehicle;
a second prediction module 300 configured to determine, based on the driving state information at the first prediction time, lane change behaviors that can be performed by the host vehicle at the first prediction time, and predict, for each lane change behavior that can be performed by the host vehicle at the first prediction time, the driving state information at a second prediction time after the lane change behavior of the host vehicle has been performed;
the first determining module 400 is configured to determine lane change decision information of the vehicle according to the driving state information at the current time, lane change behaviors that can be performed by the vehicle at the current time, the driving state information of the vehicle at each prediction time, and the lane change behaviors that can be performed by the vehicle at each prediction time.
Optionally, the number of the predicted time is N, where N is a positive integer greater than 2, the first predicted time is initially the closest predicted time to the current time among the N predicted times, and the second predicted time is the closest predicted time to the first predicted time among the N predicted times;
the device further comprises:
and a triggering module configured to, after the second prediction module predicts the travel state information of the host vehicle at a second prediction time, if a prediction time is left after the second prediction time, set the second prediction time as a new first prediction time, and trigger the second prediction model to determine a lane change behavior that the host vehicle can perform at the first prediction time, based on the travel state information at the first prediction time, and predict the travel state information at the second prediction time after the lane change behavior of the host vehicle is performed for each lane change behavior that the host vehicle can perform at the first prediction time, until the travel state information of the host vehicle at each prediction time is predicted.
Optionally, the first determining module includes:
a first processing sub-module, configured to obtain a target tree according to the driving state information at the current time, a lane change behavior that can be performed by a host vehicle at the current time, the driving state information at each prediction time, and a lane change behavior that can be performed by a host vehicle at each prediction time, where a root node of the target tree corresponds to the current time and represents the driving state information of the host vehicle at the current time, the target tree includes M layers of sub-nodes, each layer of sub-node corresponds to a prediction time, where M is a total number of prediction times; each child node with the same father node corresponds to each lane change behavior which can be performed by the vehicle at the time corresponding to the layer where the father node is located, and represents the driving state information of the vehicle at the time corresponding to the layer where the child node is located after the vehicle performs the corresponding lane change behavior;
the first determining submodule is used for determining a target parameter of each node path in the target tree, and the size of the target parameter is used for distinguishing the advantages and disadvantages among the node paths;
the second determining submodule is used for determining an optimal node path according to the target parameter;
and the third determining submodule is used for determining lane change decision information of the vehicle according to the optimal node path.
Optionally, the first determining sub-module includes:
a fourth determining submodule, configured to determine a parameter of each node in the target tree according to one or more of the following information items: lane change behavior information, lane change state information, result information after predicting lane change behavior, and continuous information between the lane change behavior of the current node and the lane change behavior of a father node of the current node;
and the fifth determining submodule is used for determining the weighted sum of the parameters of each node in each node path as the target parameter of the node path.
Optionally, the first prediction module comprises:
a sixth determining submodule, configured to predict a motion trajectory of a peripheral obstacle through a trajectory prediction model according to the information about the environment around the host vehicle at the current time, and determine a lane change behavior that the host vehicle can perform at the current time through a lane change behavior determination model according to the motion trajectory and the state information of the host vehicle at the current time;
the second prediction module comprises:
and a seventh determining submodule, configured to predict, according to the information about the environment around the host vehicle at the first prediction time, a motion trajectory of a peripheral obstacle through the trajectory prediction model, and determine, according to the motion trajectory and the host vehicle state information at the first prediction time, a lane change behavior that the host vehicle can perform at the first prediction time through the lane change behavior determination model.
Optionally, the apparatus further comprises:
the second determining module is used for determining the road condition information of the road section where the vehicle is located at the current moment; if the vehicle is in a congested road section at the current moment, the track prediction model is a first track prediction model, and the lane change behavior determination model is a first lane change behavior determination model; and if the vehicle is located at the highway section at the current moment, the track prediction model is a second track prediction model, and the lane change behavior determination model is a second lane change behavior determination model, wherein the second track prediction model is different from the first track prediction model, and the second lane change behavior determination model is different from the first lane change behavior determination model.
Optionally, the first trajectory prediction model is implemented based on a mixed density network, and the first lane change behavior determination model is implemented based on a convolutional neural network; the second trajectory prediction model is realized based on a long-term and short-term memory network, and the second lane change behavior determination model is realized based on a Bayesian network.
With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
FIG. 6 is a block diagram illustrating a vehicle 700 according to an exemplary embodiment. The vehicle can be an unmanned vehicle, and the unmanned vehicle can be applied to scenes such as logistics, distribution, takeaway and the like; the vehicle may also be any vehicle in an automatic driving mode, for example, when a driver switches a driving mode to the automatic driving mode, during the automatic driving, lane change decision information during the automatic driving may also be obtained by the method for obtaining lane change decision information of the present disclosure. As shown in fig. 6, the vehicle 700 may include: a processor 701 and a memory 702. The vehicle 700 may also include one or more of a multimedia component 703, an input/output (I/O) interface 704, and a communication component 705.
The processor 701 is configured to control the overall operation of the vehicle 700 to complete all or part of the steps in the method for obtaining lane change decision information. The memory 702 is used to store various types of data to support operation at the vehicle 700, such as instructions for any application or method operating on the vehicle 700, as well as application-related data, such as contact data, messaging, pictures, audio, video, and the like. The Memory 702 may be implemented by any type of volatile or non-volatile Memory device or combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk, or optical disk. The multimedia components 703 may include screen and audio components. Wherein the screen may be, for example, a touch screen and the audio component is used for outputting and/or inputting audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signal may further be stored in the memory 702 or transmitted through the communication component 705. The audio assembly also includes at least one speaker for outputting audio signals. The I/O interface 704 provides an interface between the processor 701 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 705 is used for wired or wireless communication between the vehicle 700 and other devices. Wireless communication, such as Wi-Fi, bluetooth, Near Field Communication (NFC), 2G, 3G, 4G, NB-IOT, eMTC, or other 5G, etc., or a combination of one or more of them, which is not limited herein. The corresponding communication component 705 may thus include: Wi-Fi module, Bluetooth module, NFC module, etc.
In an exemplary embodiment, the vehicle 700 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components for performing the above-described method of obtaining lane change decision information.
In another exemplary embodiment, a computer readable storage medium comprising program instructions which, when executed by a processor, implement the steps of the above-described method of obtaining lane change decision information is also provided. For example, the computer readable storage medium may be the memory 702 described above including program instructions executable by the processor 701 of the vehicle 700 to perform the method described above for obtaining lane change decision information.
FIG. 7 is a block diagram illustrating a vehicle 1900 according to an exemplary embodiment. Referring to fig. 7, a vehicle 1900 includes a processor 1922, which may be one or more in number, and a memory 1932 for storing computer programs executable by the processor 1922. The computer program stored in memory 1932 may include one or more modules that each correspond to a set of instructions. Further, the processor 1922 may be configured to execute the computer program to perform the above-described method of obtaining lane change decision information.
Additionally, the vehicle 1900 may also include a power component 1926 and a communication component 1950, the power component 1926 may be configured to perform power management of the vehicle 1900, and the communication component 1950 may be configured to enable communication, e.g., wired or wireless communication, of the vehicle 1900. In addition, the vehicle 1900 may also include an input/output (I/O) interface 1958. The vehicle 1900 may operate based on an operating system stored in memory 1932, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, and the like.
In another exemplary embodiment, a computer readable storage medium comprising program instructions which, when executed by a processor, implement the steps of the above-described method of obtaining lane change decision information is also provided. For example, the computer readable storage medium may be the memory 1932 described above that includes program instructions executable by the processor 1922 of the vehicle 1900 to perform the method described above for obtaining lane-change decision information.
In another exemplary embodiment, a computer program product is also provided, which comprises a computer program executable by a programmable apparatus, the computer program having code portions for performing the above-mentioned method of obtaining lane change decision information when executed by the programmable apparatus.
The preferred embodiments of the present disclosure are described in detail with reference to the accompanying drawings, however, the present disclosure is not limited to the specific details of the above embodiments, and various simple modifications may be made to the technical solution of the present disclosure within the technical idea of the present disclosure, and these simple modifications all belong to the protection scope of the present disclosure.
It should be noted that the various features described in the above embodiments may be combined in any suitable manner without departing from the scope of the invention. In order to avoid unnecessary repetition, various possible combinations will not be separately described in this disclosure.
In addition, any combination of various embodiments of the present disclosure may be made, and the same should be considered as the disclosure of the present disclosure, as long as it does not depart from the spirit of the present disclosure.

Claims (10)

1. A method for obtaining lane change decision information, the method comprising:
acquiring running state information of the vehicle at the current moment, wherein the running state information comprises the vehicle state information and the surrounding environment information of the vehicle;
determining lane change behaviors which can be performed by the vehicle at the current time according to the running state information at the current time, and predicting the running state information at a first prediction time after the lane change behaviors are performed by the vehicle for each lane change behavior which can be performed by the vehicle at the current time;
determining lane change behaviors that the host vehicle can perform at the first prediction time according to the driving state information at the first prediction time, and predicting the driving state information at a second prediction time after the host vehicle performs the lane change behaviors for each lane change behavior that the host vehicle can perform at the first prediction time;
and determining lane change decision information of the vehicle according to the running state information at the current time, the lane change behaviors which can be carried out by the vehicle at the current time, the running state information of the vehicle at each prediction time and the lane change behaviors which can be carried out by the vehicle at each prediction time.
2. The method according to claim 1, wherein the predicted time is N, where N is a positive integer greater than 2, the first predicted time is initially the closest predicted time to the current time among the N predicted times, and the second predicted time is the closest predicted time to the first predicted time among the N predicted times;
after predicting the travel state information of the host vehicle at the second prediction time, the method further includes:
and a step of, when there is a predicted time after the second predicted time, setting the second predicted time as a new first predicted time, and executing the step of determining a lane change behavior that can be performed by the host vehicle at the first predicted time again based on the travel state information at the first predicted time, and predicting the travel state information at the second predicted time after the lane change behavior of the host vehicle is performed for each lane change behavior that can be performed by the host vehicle at the first predicted time until the travel state information at each predicted time is predicted.
3. The method of claim 1, wherein said determining lane change decision information for the host vehicle based on the driving status information at the current time, the lane change behavior that may be performed by the host vehicle at the current time, the driving status information at each predicted time, and the lane change behavior that may be performed by the host vehicle at each predicted time comprises:
obtaining a target tree according to the running state information at the current moment, the lane change behavior which can be performed by the vehicle at the current moment, the running state information of the vehicle at each prediction moment and the lane change behavior which can be performed by the vehicle at each prediction moment, wherein a root node of the target tree corresponds to the current moment and represents the running state information of the vehicle at the current moment, the target tree comprises M layers of sub-nodes, each layer of sub-node corresponds to one prediction moment, and M is the total number of the prediction moments; each child node with the same father node corresponds to each lane change behavior which can be performed by the vehicle at the time corresponding to the layer where the father node is located, and represents the driving state information of the vehicle at the time corresponding to the layer where the child node is located after the vehicle performs the corresponding lane change behavior;
determining a target parameter of each node path in the target tree, wherein the size of the target parameter is used for distinguishing the advantages and disadvantages among the node paths;
determining an optimal node path according to the target parameters;
and determining lane change decision information of the vehicle according to the optimal node path.
4. The method of claim 3, wherein determining the target parameter for each node path in the target tree comprises:
determining parameters for each node in the target tree from one or more of the following items of information: lane change behavior information, lane change state information, result information after predicting lane change behavior, and continuous information between the lane change behavior of the current node and the lane change behavior of a father node of the current node;
for each node path, determining a weighted sum of the parameters of each node in the node path as a target parameter of the node path.
5. The method of claim 1, wherein determining lane change behavior that the host vehicle may perform at the current time based on the driving status information at the current time comprises:
predicting the motion trail of surrounding obstacles through a trail prediction model according to the surrounding environment information of the vehicle at the current moment, and determining lane change behaviors which can be performed by the vehicle at the current moment through a lane change behavior determination model according to the motion trail and the state information of the vehicle at the current moment;
the determining, according to the driving state information at the first prediction time, a lane change behavior that can be performed by the host vehicle at the first prediction time includes:
and predicting the motion trail of the surrounding obstacle through the trail prediction model according to the surrounding environment information of the vehicle at the first prediction time, and determining lane change behaviors which can be performed by the vehicle at the first prediction time through the lane change behavior determination model according to the motion trail and the vehicle state information at the first prediction time.
6. The method of claim 5, further comprising:
determining road condition information of a road section where the vehicle is located at the current moment;
if the vehicle is in a congested road section at the current moment, the track prediction model is a first track prediction model, and the lane change behavior determination model is a first lane change behavior determination model;
and if the vehicle is located at the highway section at the current moment, the track prediction model is a second track prediction model, and the lane change behavior determination model is a second lane change behavior determination model, wherein the second track prediction model is different from the first track prediction model, and the second lane change behavior determination model is different from the first lane change behavior determination model.
7. The method of claim 6, wherein the first trajectory prediction model is implemented based on a mixed density network and the first lane change behavior determination model is implemented based on a convolutional neural network; the second trajectory prediction model is realized based on a long-term and short-term memory network, and the second lane change behavior determination model is realized based on a Bayesian network.
8. An apparatus for obtaining lane change decision information, the apparatus comprising:
the device comprises a first acquisition module, a second acquisition module and a control module, wherein the first acquisition module is used for acquiring the driving state information of the vehicle at the current moment, and the driving state information comprises the vehicle state information and the surrounding environment information of the vehicle;
a first prediction module, configured to determine, according to the driving state information at the current time, a lane change behavior that can be performed by the host vehicle at the current time, and predict, for each lane change behavior that can be performed by the host vehicle at the current time, the driving state information at a first prediction time after the lane change behavior is performed by the host vehicle;
a second prediction module configured to determine, based on the travel state information at the first prediction time, lane change behaviors that can be performed by the host vehicle at the first prediction time, and predict, for each lane change behavior that can be performed by the host vehicle at the first prediction time, the travel state information at a second prediction time after the lane change behavior of the host vehicle has been performed;
the first determining module is used for determining lane change decision information of the vehicle according to the running state information at the current time, the lane change behaviors which can be carried out by the vehicle at the current time, the running state information of the vehicle at each prediction time and the lane change behaviors which can be carried out by the vehicle at each prediction time.
9. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 7.
10. A vehicle, characterized by comprising:
a memory having a computer program stored thereon;
a processor for executing the computer program in the memory to carry out the steps of the method of any one of claims 1 to 7.
CN201911114524.2A 2019-11-14 2019-11-14 Method and device for acquiring lane change decision information, storage medium and vehicle Withdrawn CN110908375A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911114524.2A CN110908375A (en) 2019-11-14 2019-11-14 Method and device for acquiring lane change decision information, storage medium and vehicle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911114524.2A CN110908375A (en) 2019-11-14 2019-11-14 Method and device for acquiring lane change decision information, storage medium and vehicle

Publications (1)

Publication Number Publication Date
CN110908375A true CN110908375A (en) 2020-03-24

Family

ID=69817708

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911114524.2A Withdrawn CN110908375A (en) 2019-11-14 2019-11-14 Method and device for acquiring lane change decision information, storage medium and vehicle

Country Status (1)

Country Link
CN (1) CN110908375A (en)

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111645687A (en) * 2020-06-11 2020-09-11 知行汽车科技（苏州）有限公司 Lane changing strategy determining method, device and storage medium
CN112530202A (en) * 2020-11-23 2021-03-19 中国第一汽车股份有限公司 Prediction method, device and equipment for vehicle lane change and vehicle
CN112562328A (en) * 2020-11-27 2021-03-26 腾讯科技（深圳）有限公司 Vehicle behavior prediction method and device
CN112721929A (en) * 2021-01-11 2021-04-30 成都语动未来科技有限公司 Decision-making method for lane changing behavior of automatic driving vehicle based on search technology
CN112721948A (en) * 2021-01-11 2021-04-30 成都语动未来科技有限公司 Method for realizing lane change scheduling of automatic driving automobile based on prediction and search framework
CN113077619A (en) * 2020-07-08 2021-07-06 中移(上海)信息通信科技有限公司 Method, device, equipment and storage medium for vehicle motion prediction
WO2022052856A1 (en) * 2020-09-10 2022-03-17 腾讯科技（深圳）有限公司 Vehicle-based data processing method and apparatus, computer, and storage medium
US11433923B2 (en) 2020-06-10 2022-09-06 Toyota Motor Engineering & Manufacturing North America, Inc. Systems and methods for long-term prediction of lane change maneuver
CN116499487A (en) * 2023-06-28 2023-07-28 新石器慧通（北京）科技有限公司 Vehicle path planning method, device, equipment and medium

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106364486A (en) * 2016-09-08 2017-02-01 江苏大学 Intelligent vehicle lane changing control method based on hazard analysis
CN108364467A (en) * 2018-02-12 2018-08-03 北京工业大学 A kind of traffic information prediction technique based on modified decision Tree algorithms
CN109739246A (en) * 2019-02-19 2019-05-10 百度在线网络技术（北京）有限公司 Decision-making technique, device, equipment and storage medium during a kind of changing Lane
CN109910880A (en) * 2019-03-07 2019-06-21 百度在线网络技术（北京）有限公司 Method, apparatus, storage medium and the terminal device of vehicle behavior planning
CN110020748A (en) * 2019-03-18 2019-07-16 杭州飞步科技有限公司 Trajectory predictions method, apparatus, equipment and storage medium
CN110356401A (en) * 2018-04-05 2019-10-22 北京图森未来科技有限公司 A kind of automatic driving vehicle and its lane change control method and system

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106364486A (en) * 2016-09-08 2017-02-01 江苏大学 Intelligent vehicle lane changing control method based on hazard analysis
CN108364467A (en) * 2018-02-12 2018-08-03 北京工业大学 A kind of traffic information prediction technique based on modified decision Tree algorithms
CN110356401A (en) * 2018-04-05 2019-10-22 北京图森未来科技有限公司 A kind of automatic driving vehicle and its lane change control method and system
CN109739246A (en) * 2019-02-19 2019-05-10 百度在线网络技术（北京）有限公司 Decision-making technique, device, equipment and storage medium during a kind of changing Lane
CN109910880A (en) * 2019-03-07 2019-06-21 百度在线网络技术（北京）有限公司 Method, apparatus, storage medium and the terminal device of vehicle behavior planning
CN110020748A (en) * 2019-03-18 2019-07-16 杭州飞步科技有限公司 Trajectory predictions method, apparatus, equipment and storage medium

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11433923B2 (en) 2020-06-10 2022-09-06 Toyota Motor Engineering & Manufacturing North America, Inc. Systems and methods for long-term prediction of lane change maneuver
CN111645687A (en) * 2020-06-11 2020-09-11 知行汽车科技（苏州）有限公司 Lane changing strategy determining method, device and storage medium
CN113077619A (en) * 2020-07-08 2021-07-06 中移(上海)信息通信科技有限公司 Method, device, equipment and storage medium for vehicle motion prediction
CN113077619B (en) * 2020-07-08 2021-12-07 中移(上海)信息通信科技有限公司 Method, device, equipment and storage medium for vehicle motion prediction
WO2022052856A1 (en) * 2020-09-10 2022-03-17 腾讯科技（深圳）有限公司 Vehicle-based data processing method and apparatus, computer, and storage medium
CN112530202A (en) * 2020-11-23 2021-03-19 中国第一汽车股份有限公司 Prediction method, device and equipment for vehicle lane change and vehicle
CN112562328A (en) * 2020-11-27 2021-03-26 腾讯科技（深圳）有限公司 Vehicle behavior prediction method and device
CN112721929A (en) * 2021-01-11 2021-04-30 成都语动未来科技有限公司 Decision-making method for lane changing behavior of automatic driving vehicle based on search technology
CN112721948A (en) * 2021-01-11 2021-04-30 成都语动未来科技有限公司 Method for realizing lane change scheduling of automatic driving automobile based on prediction and search framework
CN116499487A (en) * 2023-06-28 2023-07-28 新石器慧通（北京）科技有限公司 Vehicle path planning method, device, equipment and medium
CN116499487B (en) * 2023-06-28 2023-09-05 新石器慧通（北京）科技有限公司 Vehicle path planning method, device, equipment and medium

Similar Documents

Publication Publication Date Title
CN110908375A (en) Method and device for acquiring lane change decision information, storage medium and vehicle
US10739774B2 (en) Keyframe based autonomous vehicle operation
CN112840350B (en) Autonomous vehicle planning and prediction
US10739776B2 (en) Autonomous vehicle policy generation
CN110364006B (en) Machine learning enhanced vehicle merge
CN111208838B (en) Control method and device of unmanned equipment
CN110262486B (en) Unmanned equipment motion control method and device
CN112419722B (en) Traffic abnormal event detection method, traffic control method, device and medium
CN112629550B (en) Method and device for predicting obstacle track and model training
US11941509B2 (en) Method and system for determining information on an expected trajectory of an object
CN112306059B (en) Training method, control method and device for control model
CN113341941B (en) Control method and device of unmanned equipment
US20230281357A1 (en) Generating simulation environments for testing av behaviour
US20230331247A1 (en) Systems for testing and training autonomous vehicles
Ilić et al. Predicting positions and velocities of surrounding vehicles using deep neural networks
US20220326714A1 (en) Unmapped u-turn behavior prediction using machine learning
CN111123957B (en) Method and device for planning track
CN113753040A (en) Predicting road disorderly crossing behavior of weak road users
CN113246973B (en) Obstacle trajectory prediction method, obstacle trajectory prediction device, storage medium, and electronic device
CN115817466A (en) Collision risk assessment method and device
CN115047864A (en) Model training method, unmanned equipment control method and device
CN114194213A (en) Target object trajectory prediction method and device, storage medium and electronic equipment
CN114114954A (en) Control method and device for unmanned equipment
CN115116236B (en) Method and device for predicting driving behavior, storage medium and electronic device
CN118552931A (en) Track prediction model training, track prediction and track planning methods and devices

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WW01 Invention patent application withdrawn after publication

Application publication date: 20200324

WW01 Invention patent application withdrawn after publication