CN111284485B - Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium - Google Patents

Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium Download PDF

Info

Publication number
CN111284485B
CN111284485B CN201910959765.0A CN201910959765A CN111284485B CN 111284485 B CN111284485 B CN 111284485B CN 201910959765 A CN201910959765 A CN 201910959765A CN 111284485 B CN111284485 B CN 111284485B
Authority
CN
China
Prior art keywords
vehicle
obstacle
determining
obstacle vehicle
driving
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.)
Active
Application number
CN201910959765.0A
Other languages
Chinese (zh)
Other versions
CN111284485A (en
Inventor
李宇寂
尚秉旭
王洪峰
陈志新
刘洋
许朝文
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.)
FAW Group Corp
Original Assignee
FAW Group Corp
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 FAW Group Corp filed Critical FAW Group Corp
Priority to CN201910959765.0A priority Critical patent/CN111284485B/en
Publication of CN111284485A publication Critical patent/CN111284485A/en
Application granted granted Critical
Publication of CN111284485B publication Critical patent/CN111284485B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
    • B60W30/08—Active safety systems predicting or avoiding probable or impending collision or attempting to minimise its consequences
    • B60W30/095—Predicting travel path or likelihood of collision
    • B60W30/0956—Predicting travel path or likelihood of collision the prediction being responsive to traffic or environmental parameters
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
    • B60W40/02—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to ambient conditions
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W50/0097—Predicting future conditions
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W50/0098—Details of control systems ensuring comfort, safety or stability not otherwise provided for
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B60—VEHICLES IN GENERAL
    • B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
    • B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
    • B60W2050/0001—Details of the control system
    • B60W2050/0043—Signal treatments, identification of variables or parameters, parameter estimation or state estimation

Landscapes

  • Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Traffic Control Systems (AREA)
  • Control Of Driving Devices And Active Controlling Of Vehicle (AREA)

Abstract

The invention discloses a method and a device for predicting driving behaviors of an obstacle vehicle, the vehicle and a storage medium. The method comprises the following steps: obtaining relative position information of an obstacle vehicle, determining the accurate position of the obstacle vehicle based on positioning information of the vehicle and the relative position information of the obstacle vehicle, generating a vehicle motion track according to the accurate position of the obstacle vehicle within a preset time, generating an observation value sequence according to the vehicle motion track, obtaining corresponding forward probabilities of the observation value sequence through at least one preset driving action model, determining the maximum forward probability from the forward probabilities, and determining the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability. The problem that only the forward vehicle can be predicted when the driving behavior of the obstacle vehicle is predicted is solved, the effect of predicting the driving behavior of any vehicle around is achieved, and the safe and efficient driving route is better planned by accurately judging the possible driving intention of the obstacle vehicle.

Description

Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium
Technical Field
The embodiment of the invention relates to an automobile driving technology, in particular to a method and a device for predicting driving behavior of an obstacle vehicle, the vehicle and a storage medium.
Background
In a complex traffic environment, a vehicle needs to run safely and quickly, a path of the vehicle needs to be planned, and a next driving action needs to be planned dynamically according to real-time changes of surrounding vehicles.
Currently, an experienced driver can accurately predict the next behavior of the vehicle according to a series of actions of the surrounding obstacle vehicle; in the automatic driving application, the unmanned vehicle also needs to accurately plan the next action of the self vehicle according to the driving behaviors of surrounding vehicles, and the automatic driving vehicle predicts the driving behaviors of surrounding obstacle vehicles and predicts the next driving action of a forward vehicle by analyzing the driving behaviors of the forward vehicle.
The driving behavior of the front vehicle is predicted only by the automatic driving vehicle for predicting the driving behavior of the front obstacle vehicle in the driving process of the vehicle, the driving behaviors of the side and rear vehicles of the vehicle cannot be obtained, enough obstacle vehicle information cannot be provided for the unmanned vehicle, and the automatic driving vehicle cannot adapt to complex traffic road conditions.
Disclosure of Invention
The invention provides a method and a device for predicting the driving behavior of an obstacle vehicle, a vehicle and a storage medium, which are used for accurately predicting the driving behavior of peripheral obstacle vehicles.
In a first aspect, an embodiment of the present invention provides a method for predicting driving behavior of an obstacle vehicle, where the method includes:
acquiring relative position information of an obstacle vehicle;
determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle;
generating a vehicle motion track according to the accurate position of an obstacle vehicle within preset time, and generating an observation value sequence according to the vehicle motion track;
obtaining corresponding forward probabilities of the observation value sequence through at least one preset driving action model, and determining the maximum forward probability from each forward probability;
and determining the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability.
In a second aspect, an embodiment of the present invention also provides a device for predicting driving behavior of an obstacle vehicle, including:
the acquisition module is used for acquiring relative position information of the obstacle vehicle;
the positioning module is used for determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle;
the generating module is used for generating a vehicle motion track according to the accurate position of the obstacle vehicle within the preset time and generating an observation value sequence according to the vehicle motion track;
the probability determination module is used for obtaining corresponding forward probabilities from the observation value sequence through at least one preset driving action model and determining the maximum forward probability from the forward probabilities;
and the action determining module is used for determining the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability.
In a third aspect, an embodiment of the present invention further provides a vehicle, including:
one or more sensors for collecting relative position information of an obstacle vehicle;
one or more controllers;
a storage device for storing one or more programs,
when the one or more programs are executed by the one or more controllers, the one or more controllers may be caused to implement a method of predicting driving behavior of an obstacle vehicle as described in any one of the embodiments of the present invention.
In a fourth aspect, the embodiments of the present invention also provide a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements a method for predicting driving behavior of an obstacle vehicle according to any one of the embodiments of the present invention.
The invention determines the accurate position of the obstacle vehicle by acquiring the relative position information of the obstacle vehicle, combines the relative position information of the obstacle vehicle based on the positioning information of the vehicle, generates the vehicle motion track according to the accurate position of the obstacle vehicle within the preset time, generates the observation value sequence according to the vehicle motion track, obtains the corresponding forward probability by at least one preset driving action model of the observation value sequence, determines the maximum forward probability from the forward probabilities, determines the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability, solves the problem that the obstacle vehicle can only predict the forward vehicle when predicting the driving action, realizes the effect of predicting the driving action of any vehicle around, and assists the automatic driving vehicle to carry out effective driving decision by accurately judging the possible driving intention of the obstacle vehicle around, and a safe and efficient driving route is better planned.
Drawings
Fig. 1 is a flowchart of a method for predicting driving behavior of an obstacle vehicle according to a first embodiment of the present invention;
fig. 2 is a flowchart of a method for predicting driving behavior of an obstacle vehicle according to a second embodiment of the present invention;
FIG. 3 is a diagram showing an example of the relative position of the vehicle in the second embodiment of the invention;
FIG. 4 is a schematic diagram of a vehicle movement trace according to a second embodiment of the present invention;
FIG. 5 is a diagram illustrating an exemplary direction vector according to a second embodiment of the present invention;
FIG. 6 is a schematic diagram illustrating a process of determining an angle coding association table according to a second embodiment of the present invention;
FIG. 7 is a schematic diagram of direction division in the second embodiment of the present invention;
FIG. 8 is a diagram illustrating quantization rules according to a second embodiment of the present invention;
FIG. 9 is a schematic view of an observation sequence in a second embodiment of the present invention;
FIG. 10 is a diagram showing an example of a classic structure of an HMM model in the second embodiment of the present invention;
FIG. 11 is a diagram illustrating an example of an HMM model training process for a single driving action in accordance with a second embodiment of the present invention;
FIG. 12 is a diagram illustrating an example of dividing an expressway scene in a second embodiment of the present invention;
fig. 13 is an exemplary diagram of urban road scene division in the second embodiment of the present invention;
FIG. 14 is a schematic diagram of a generation process of a preset threshold model according to a second embodiment of the present invention;
fig. 15 is a flowchart of the obstacle driving action prediction in the second embodiment of the invention;
fig. 16 is an overall configuration diagram of a driving action of a vehicle predicted to be hindered according to the second embodiment of the present invention;
fig. 17 is a structural diagram of a driving behavior prediction apparatus for an obstacle vehicle in a third embodiment of the present invention;
fig. 18 is a schematic structural diagram of a vehicle in a fourth embodiment of the invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Example one
Fig. 1 is a flowchart of a method for predicting driving behavior of an obstacle vehicle according to an embodiment of the present invention, where the present embodiment is applicable to a case of predicting driving behavior of an obstacle vehicle, and the method may be executed by an obstacle vehicle driving behavior prediction device, and specifically includes the following steps:
and step 110, acquiring relative position information of the obstacle vehicle.
The obstacle vehicle can be specifically understood as any vehicle running within a certain distance on the current road; the relative position information may be position information of the obstacle vehicle from the own vehicle.
Specifically, the relative position information of the obstacle vehicle can be obtained by a laser radar, a camera and/or a millimeter wave radar, the relative position information of other obstacle vehicles around the vehicle can be obtained by at least one sensor, and the more accurate and robust relative position of the obstacle vehicle can be obtained by a multi-sensor fusion technology; the relative position information may be distance and angle information.
The obstacle vehicle is any vehicle within a certain distance and can be a front vehicle, a rear vehicle and/or a side vehicle, the problem that only the front vehicle can be predicted in the driving behavior prediction of the obstacle vehicle is solved, the possible driving intention of the surrounding obstacle vehicle is accurately judged, the automatic driving vehicle is assisted to carry out effective driving decision, and a safe and efficient driving route is better planned.
And step 120, determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle.
The positioning information can be specifically understood as specific position information of the vehicle in the global positioning system; the accurate position is the specific position information of the obstacle vehicle in the global positioning system.
Specifically, the positioning information may be obtained by a navigation positioning system, where the navigation positioning system may be at least one of the following: a Beidou satellite system, a GPS system, a GLONASS system, a GALILEO system, a QZSS system and/or an SBAS system. The manner of determining the exact position of the obstacle vehicle may be: the absolute position and the course angle of the vehicle in the world coordinate system can be obtained by the vehicle through high-precision positioning information, the relative position of the target obstacle vehicle in the vehicle coordinate system can be obtained through the sensor, and the absolute coordinate of the target obstacle vehicle in the world coordinate system is finally determined.
And step 130, generating a vehicle motion track according to the accurate position of the obstacle vehicle within the preset time, and generating an observation value sequence according to the vehicle motion track.
The preset time can be specifically understood as a period of time preset according to the requirement and used for acquiring a plurality of driving actions of the obstacle vehicle; the observation sequence is a string of numbers from 1 to 16.
Specifically, the observation value sequence may be generated by processing a vehicle motion trajectory according to a preset coding rule and then generating a quantized code.
And 140, obtaining corresponding forward probabilities of the observation value sequence through at least one preset driving action model, and determining the maximum forward probability from each forward probability.
The preset driving action model can be specifically understood as a driving action model trained according to driving actions, and a plurality of driving action models can be trained according to actual driving conditions and used for calculating and obtaining corresponding forward probabilities according to the observation value sequence.
Specifically, the maximum forward probability may be determined by comparing forward probabilities obtained by different preset driving action models.
And 150, determining the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability.
Specifically, the manner of determining the driving action of the obstacle vehicle may be determined by determining a driving action model corresponding to the maximum forward probability.
The invention determines the accurate position of the obstacle vehicle by acquiring the relative position information of the obstacle vehicle, combines the relative position information of the obstacle vehicle based on the positioning information of the vehicle, generates the vehicle motion track according to the accurate position of the obstacle vehicle within the preset time, generates the observation value sequence according to the vehicle motion track, obtains the corresponding forward probability by at least one preset driving action model of the observation value sequence, determines the maximum forward probability from the forward probabilities, determines the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability, solves the problem that the obstacle vehicle can only predict the forward vehicle when predicting the driving action, realizes the effect of predicting the driving action of any vehicle around, and assists the automatic driving vehicle to carry out effective driving decision by accurately judging the possible driving intention of the obstacle vehicle, and a safe and efficient driving route is better planned.
Example two
Fig. 2 is a flowchart of a method for predicting driving behavior of an obstacle vehicle according to a second embodiment of the present invention. The technical scheme of the embodiment is further refined on the basis of the technical scheme, and specifically mainly comprises the following steps:
and step 210, acquiring relative position information of the obstacle vehicle.
And step 220, determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle.
Specifically, fig. 3 shows an example of a relative position relationship between vehicles, and the method for determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle may be: the absolute position and the course angle (X) of the vehicle under the world coordinate system can be obtained by the vehicle through high-precision positioning informationω,Yω) And beta, the relative position (X) of the target vehicle in the coordinate system of the vehicle can be obtained by the sensorυ,Yυ) The final absolute coordinates (X) of the target vehicle in the world coordinate systemf,Yf) Can be obtained by the following formula:
Figure BDA0002228533610000071
and step 230, generating a vehicle motion track according to the accurate position of the obstacle vehicle within the preset time.
Specifically, fig. 4 shows a schematic diagram of a motion trajectory of a vehicle. The driving of the vehicle is continuous in time series, and the vehicle normally runs along a given driving intention and does not move disorderly in a short period of time, so that the driving action of the vehicle in a historical period is recognized at the current moment, and the driving intention of the vehicle in a future period can be predicted.
And 240, determining a direction vector from each moment position to a corresponding later moment position of the obstacle vehicle within a preset time according to the vehicle motion track.
Specifically, fig. 5 illustrates an example of a direction vector formed by a vehicle from a respective time position to a corresponding subsequent time position over a period of time.
And step 250, determining the angle value corresponding to each direction vector, and searching an angle coding association table to obtain the numerical label corresponding to each angle value.
The angle code association table may be specifically understood as a table formed by one-to-one correspondence of angle values and numerical labels.
Specifically, the manner of searching the angle coding association table to obtain the digital labels corresponding to the angle values may be to determine the angle values according to the direction vectors, and search and match the digital labels corresponding to the angle values according to the angle values and the angle coding association table.
Further, fig. 6 shows a schematic diagram of a determination process of the angle coding association table. As shown in fig. 6, the determination method of the angle coding association table specifically includes the following steps:
step 2501, equally dividing the circumferential direction into 16 parts, and carrying out degree labeling in a counterclockwise direction from the positive direction of the abscissa to obtain degree values of each part.
Specifically, the circumference is equally divided into 16 portions each of 22.5 °, and degrees are numbered starting with the positive direction of the abscissa as 0 °, and a schematic diagram of direction division is given in fig. 7.
Step 2502, equally dividing the plane rectangular coordinate system into 16 parts, and labeling the parts according to numbers 1-16 clockwise in the negative direction of the abscissa to obtain the numerical labels of the parts.
Specifically, the rectangular plane coordinate system is divided equally, the obtained number labels of the respective shares are quantization rules, starting from the negative direction of the abscissa, the number labels are performed on each share, the first share is 1, the second share is 2 …, and the schematic diagram of the quantization rules is given in fig. 8.
Step 2503, associating each degree value with a numerical label according to positions to form an angle coding association table.
Specifically, each degree value interval corresponds to a number label, the degree value interval and the number label of the corresponding position can be determined according to the direction division schematic diagram and the quantization rule schematic diagram, and the degree value interval and the number label can correspond to 8 when the degree is more than or equal to 0 degrees and less than 22.5 degrees.
And step 260, forming an observation value sequence of the vehicle track sequence based on each digital label.
Specifically, fig. 9 shows an observation value sequence diagram, different driving actions may form different vehicle driving tracks, different direction feature vectors may be formed after quantization coding, and the driving actions may be finally distinguished by distinguishing the direction feature vectors.
And 270, determining the selectable preset driving action models of the vehicle under different scenes according to the accurate position of the obstacle vehicle.
The different scenes can be specifically understood as different road types where the vehicle is located, wherein the road types comprise an expressway and an urban road and are used for more specifically limiting the output result of the vehicle needing to predict the driving behavior.
Specifically, scene information of the vehicle, such as a high-speed scene or an urban road scene, can be obtained according to the navigation map, and the scene can be further refined: the high-speed scene is divided into a common road scene and an upper and lower high-speed ramp scene; urban road scenes can be divided into common road scenes and intersection scenes, and intersections can be divided into different scenes such as crossroads, T-intersections and the like.
Specifically, the preset driving action model is used for judging the probability of possible driving action of the vehicle through a historical track sequence of the vehicle in a period of time. Alternatively, the statistical processing is performed by a Hidden Markov Model (HMM). The HMM model is used for carrying out statistical analysis processing on different driving actions, the feature vectors of the different driving actions are required to be obtained firstly, and then the models corresponding to the different driving actions are obtained through offline model training. In the actual vehicle running process, a track sequence generated by an obstacle vehicle for a period of time is quantized into a feature vector, the feature vector is input into a trained model, and a final driving action judgment result can be obtained after judgment is carried out through a BAUM-WELCH algorithm. Designing an HMM model λ ═ (a, B, pi) as a classical left-right structure, fig. 10 gives an exemplary diagram of the classical structure of the HMM model, and as shown in fig. 10, specific parameters of the model are as follows: the number of the state sets is N (set to be 3), the number of the observation sets is M (set to be 16), the state transition matrix A is a matrix of N x N, the output probability distribution matrix B is a matrix of N x M, the initialization vector pi is a vector of 1 x N, and specific parameters can be adjusted according to different requirements. The specific process of the training stage is as follows: and forming a corresponding direction feature vector by each action sequence in the training set, training by using an HMM B-W algorithm to obtain a corresponding model, wherein each model has a plurality of groups of action sequences, and averaging all the models of the same action to obtain a final HMM model corresponding to the action, such as a straight-line HMM model. FIG. 11 is a diagram illustrating an example of an HMM model training process for a single driving action.
The driving of the vehicle is continuous in time series, and the vehicle normally runs along a given driving intention and does not move disorderly in a short period of time, so that the driving action of the vehicle in a historical period is recognized at the current moment, and the driving intention of the vehicle in a future period can be predicted. The method for statistically judging by applying the hidden Markov model can judge the driving intention of the vehicle without completely executing the action because the probability of different possible driving actions is calculated, for example, the probability that the vehicle can judge future turning in the early stage of turning is far greater than the probability of continuing straight running by the model.
Further, the preset driving action model includes: the system comprises a straight-going model, a left lane-changing model, a right lane-changing model, a left turning model and a right turning model.
When the vehicle runs to an actual working condition, possible driving behaviors of the vehicle can be determined in more detail according to specific subdivision scenes, for example, in a ramp scene on a highway, the vehicle can only change lanes in a straight direction and in a left direction; in a down-ramp scene, vehicles can only change lanes in a straight line and a right line, and an example diagram for dividing a highway scene is shown in fig. 12; for intersection scenes of common roads, such as non-intersections, e.g., t-intersections or intersections prohibited from turning left, the possible driving directions of the vehicle are only straight and turning right, and fig. 13 shows an example of dividing urban road scenes. According to the specific scene of the current vehicle, several possible actions of the vehicle can be determined, the prediction of the impossible actions is omitted, the prediction process is simplified, and the prediction accuracy is improved.
By applying the statistical processing method, different driving habits can be better classified, so that individual differences of different drivers are eliminated, for example, the driving habits with different turning radii can be judged to be left-turning or right-turning, so that the prediction of the driving behavior has better generalization and robustness. The prediction aiming at different traffic scenes can ensure that the prediction process has better pertinence on one hand, and can also simplify the prediction process on the other hand, and eliminate unnecessary false recognition, for example, in a high-speed scene, an obstacle vehicle cannot turn left or right, so that the left or right turning does not need to be considered in the output prediction result, the calculation complexity of a program is greatly reduced, and the false recognition is eliminated.
And step 280, obtaining corresponding forward probabilities of the observation value sequence through at least one preset driving action model, and determining the maximum forward probability from each forward probability.
Step 290, determining the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability.
The determination result is the driving action type of the vehicle, and the driving action type can be straight running, left lane changing, right lane changing, left turning, right turning or random action.
Specifically, the driving action of the vehicle is determined according to the driving action model corresponding to the maximum forward probability. For example, if the driving action model corresponding to the maximum forward probability is a straight-ahead model, the vehicle driving action is straight-ahead.
Further, the present embodiment may specifically optimize "determining the driving action of the obstacle vehicle according to the determination result of the maximum forward probability" to: when the maximum forward probability is within a preset threshold model, determining that the driving action of the obstacle vehicle is a random action; otherwise, determining the driving action of the obstacle vehicle as the driving action corresponding to the maximum forward probability.
The preset threshold model can be specifically understood as being generated by combining common information of five driving action models and is used for distinguishing five driving actions from random actions; the random action may be a small left-right sway of the vehicle trajectory caused by the continuous small correction of the steering wheel by the novice driver.
Specifically, by comparing the maximum forward probability with a preset threshold model, if the maximum forward probability is within the preset threshold model, the driving action of the obstacle vehicle is a random action; and if not, the driving action of the obstacle vehicle is the driving action corresponding to the maximum forward probability.
Determining the maximum value of all the forward probabilities, comparing the maximum value with a preset threshold model, and if the maximum value is larger than the preset threshold model, determining that the driving behavior of the obstacle vehicle is the driving action corresponding to the maximum forward probability; if the value is smaller than the preset threshold model, the action is irregular random action, which is probably caused by irregular driving behavior of the driver.
Exemplarily, fig. 14 shows a schematic diagram of a generation process of a preset threshold model, which may be generated according to common information of five types of driving action models; FIG. 15 is a flow chart illustrating a method for predicting an obstacle driving action, where the obstacle driving action prediction may be to calculate a driving behavior probability according to an input obstacle vehicle observation sequence, and predict a driving behavior of an obstacle vehicle in combination with a threshold model; fig. 16 shows an overall structure diagram for predicting the driving action of the obstacle vehicle, the relative position information of the obstacle vehicle is obtained through various acquisition modes, the absolute position information of the obstacle vehicle is determined by combining the absolute position information of the vehicle, and the driving action of the obstacle vehicle is predicted by combining the track sequence of the obstacle vehicle and a trained driving action model.
The driving behavior of the obstacle vehicle within a period of time is statistically judged by applying a hidden Markov model, and the probability of different driving actions is output, so that the driving intention of the obstacle vehicle can be predicted: straight, left lane change, right lane change, left turn, right turn or random motion.
The invention determines the accurate position of the obstacle vehicle by acquiring the relative position information of the obstacle vehicle, combines the relative position information of the obstacle vehicle based on the positioning information of the vehicle, generates the vehicle motion track according to the accurate position of the obstacle vehicle within the preset time, generates the observation value sequence according to the vehicle motion track, obtains the corresponding forward probability by at least one preset driving action model of the observation value sequence, determines the maximum forward probability from the forward probabilities, determines the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability, solves the problem that the obstacle vehicle can only predict the forward vehicle when predicting the driving action, realizes the effect of predicting the driving action of any vehicle around, and assists the automatic driving vehicle to carry out effective driving decision by accurately judging the possible driving intention of the obstacle vehicle around, and a safe and efficient driving route is better planned. By applying a statistical processing method when a driving action model is preset, different driving habits can be better classified, so that individual differences of different drivers are eliminated, and the prediction of driving behaviors has better generalization and robustness. The prediction is carried out aiming at different traffic scenes, so that the prediction process has better pertinence, and the prediction process can be simplified and unnecessary false recognition is eliminated, thereby greatly reducing the calculation complexity of the program and eliminating the false recognition.
EXAMPLE III
Fig. 17 is a configuration diagram of a device for predicting driving behavior of an obstacle vehicle according to a third embodiment of the present invention, the device including: an acquisition module 31 and a positioning module 32, a generation module 33, a probability determination module 34 and an action determination module 35.
The acquiring module 31 is used for acquiring relative position information of the obstacle vehicle; the positioning module 32 is used for determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle; the generating module 33 is configured to generate a vehicle motion trajectory according to an accurate position of a vehicle obstructed in a preset time, and generate an observation value sequence according to the vehicle motion trajectory; a probability determining module 34, configured to obtain corresponding forward probabilities from the observation value sequence through at least one preset driving action model, and determine a maximum forward probability from each of the forward probabilities; and the action determining module 35 is configured to determine the driving action of the obstacle vehicle according to the determination result of the maximum forward probability.
According to the method, the relative position information of the obstacle vehicle is obtained through the obtaining module, the accurate position of the obstacle vehicle is determined through the positioning module, the problem that only forward vehicles can be predicted when the driving behaviors of the obstacle vehicle are predicted is solved, the effect of predicting the driving behaviors of any vehicle around is achieved, the observation value sequence is generated through the generating module according to the motion track of the vehicle, the maximum forward probability is determined according to the probability determining module, and finally the driving actions of the obstacle vehicle are determined according to the action determining module. The method has the advantages that the possible driving intention of the obstacle vehicle is accurately judged, the automatic driving vehicle is assisted to carry out effective driving decision, and a safe and efficient driving route is better planned.
Optionally, the preset driving action model includes: the system comprises a straight-going model, a left lane-changing model, a right lane-changing model, a left turning model and a right turning model.
Further, the generating module 33 includes:
the track generating unit is used for generating a vehicle motion track according to the accurate position of the obstacle vehicle within the preset time;
and the sequence generating unit is used for quantizing the vehicle motion track into codes through a preset coding rule to form an observation value sequence.
Further, the sequence generating unit is specifically configured to: determining a direction vector of the obstacle vehicle from each moment position to a corresponding later moment position within a preset time according to the vehicle motion track; determining an angle value corresponding to each direction vector, and searching an angle coding association table to obtain a digital label corresponding to each angle value; and forming an observation value sequence of the vehicle track sequence based on each digital label.
Further, the apparatus further comprises: the association table determining module is specifically used for equally dividing the circumferential direction into 16 parts, and carrying out degree labeling counterclockwise from the positive direction of the abscissa to obtain degree values of each part; dividing the rectangular plane coordinate system into 16 parts equally, and labeling the parts clockwise according to numbers 1-16 in the negative direction of the abscissa to obtain the numerical labels of the parts; and associating each degree value with a numerical label according to positions to form an angle coding association table.
In the technical solution of the above embodiment, the apparatus further includes:
and the model determining module is used for obtaining different forward probabilities through different preset driving action models in the observed value sequence, and determining the selectable preset driving action models of the vehicle under different scenes according to the accurate position of the obstacle vehicle before determining the maximum forward probability.
Further, the action determining module 35 is specifically configured to: when the maximum forward probability is within a preset threshold model, determining that the driving action of the obstacle vehicle is a random action; otherwise, determining the driving action of the obstacle vehicle as the driving action corresponding to the maximum forward probability.
The device for predicting the driving behavior of the obstacle vehicle, provided by the embodiment of the invention, can execute the method for predicting the driving behavior of the obstacle vehicle, provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
Example four
Fig. 18 is a schematic structural diagram of a vehicle according to a fourth embodiment of the present invention, as shown in fig. 18, the vehicle includes a sensor 40, a controller 41, a memory 42, an input device 43, and an output device 44; the number of sensors 40 and the number of controllers 41 in the vehicle may be one or more, and one sensor 40 and one controller 41 are exemplified in fig. 17; the sensors 40, the controller 41, the memory 42, the input device 43, and the output device 44 in the vehicle may be connected by a bus or other means, and the bus connection is exemplified in fig. 18.
And a sensor 40 for collecting relative position information of the obstacle vehicle.
The memory 42 serves as a computer-readable storage medium that can be used to store software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the obstacle vehicle driving behavior prediction method in the embodiment of the present invention (for example, the acquisition module 31 and the positioning module 32, the generation module 33, the probability determination module 34, and the action determination module 35 in the obstacle vehicle driving behavior prediction apparatus). The controller 41 executes various functional applications and data processing of the vehicle, that is, implements the above-described obstacle vehicle driving behavior prediction method, by executing software programs, instructions, and modules stored in the memory 42.
The memory 42 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal, and the like. Further, the memory 42 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, the memory 42 may further include memory remotely located from the controller 41, which may be connected to the vehicle over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 43 may be used to receive input numeric or character information and generate key signal inputs related to user settings and function control of the cloud platform. The output device 44 may include a display device such as a display screen.
EXAMPLE five
Embodiments of the present invention also provide a storage medium containing computer-executable instructions which, when executed by a computer processor, perform a method of predicting driving behavior of an obstacle vehicle, the method comprising:
acquiring relative position information of an obstacle vehicle;
determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle;
generating a vehicle motion track according to the accurate position of an obstacle vehicle within preset time, and generating an observation value sequence according to the vehicle motion track;
obtaining corresponding forward probabilities of the observation value sequence through at least one preset driving action model, and determining the maximum forward probability from each forward probability;
and determining the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability.
Of course, the storage medium containing the computer-executable instructions provided by the embodiments of the present invention is not limited to the method operations described above, and may also perform related operations in the method for predicting driving behavior of an obstacle vehicle provided by any embodiment of the present invention.
From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods according to the embodiments of the present invention.
It should be noted that, in the embodiment of the device for predicting driving behavior of an obstacle vehicle, the included units and modules are only divided according to the functional logic, but are not limited to the above division as long as the corresponding functions can be realized; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims (7)

1. A method of predicting driving behavior of a barrier vehicle, comprising:
acquiring relative position information of an obstacle vehicle;
determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle;
generating a vehicle motion track according to the accurate position of an obstacle vehicle within preset time, and generating an observation value sequence according to the vehicle motion track;
obtaining corresponding forward probabilities of the observation value sequence through at least one preset driving action model, and determining the maximum forward probability from each forward probability;
determining the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability;
generating an observation value sequence according to the vehicle motion track comprises the following steps:
quantizing the vehicle motion track into codes through a preset coding rule to form an observation value sequence;
quantizing the vehicle motion trajectory into codes through a preset coding rule to form an observation value sequence, and the method comprises the following steps:
determining a direction vector of the obstacle vehicle from each moment position to a corresponding later moment position within a preset time according to the vehicle motion track;
determining an angle value corresponding to each direction vector, and searching an angle coding association table to obtain a digital label corresponding to each angle value;
forming an observed value sequence of the vehicle motion track based on each digital label;
the angle code association table is determined by the following steps:
equally dividing the circumferential direction into 16 parts, and carrying out degree labeling counterclockwise from the positive direction of the abscissa to obtain degree values of all the parts;
dividing the rectangular plane coordinate system into 16 parts equally, and labeling the parts clockwise according to numbers 1-16 in the negative direction of the abscissa to obtain the numerical labels of the parts;
and associating each degree value with a numerical label according to positions to form an angle coding association table.
2. The method of claim 1, wherein the preset driving maneuver model comprises:
the system comprises a straight-going model, a left lane-changing model, a right lane-changing model, a left turning model and a right turning model.
3. The method of claim 2, wherein before determining the maximum forward probability after obtaining different forward probabilities from the observation sequence through different preset driving action models, further comprising:
and determining the selectable preset driving action models of the vehicle under different scenes according to the accurate position of the obstacle vehicle.
4. The method according to claim 1, wherein the determining the driving action of the obstacle vehicle according to the determination of the maximum forward probability includes:
when the maximum forward probability is within a preset threshold model, determining that the driving action of the obstacle vehicle is a random action; otherwise, determining the driving action of the obstacle vehicle as the driving action corresponding to the maximum forward probability.
5. A device for predicting driving behavior of an obstacle vehicle, characterized by comprising:
the acquisition module is used for acquiring relative position information of the obstacle vehicle;
the positioning module is used for determining the accurate position of the obstacle vehicle based on the positioning information of the vehicle and the relative position information of the obstacle vehicle;
the generating module is used for generating a vehicle motion track according to the accurate position of the obstacle vehicle within the preset time and generating an observation value sequence according to the vehicle motion track;
the sequence generating unit is used for quantizing the vehicle motion track into codes through a preset coding rule to form an observation value sequence;
a sequence generation unit, specifically configured to:
determining a direction vector of the obstacle vehicle from each moment position to a corresponding later moment position within a preset time according to the vehicle motion track;
determining an angle value corresponding to each direction vector, and searching an angle coding association table to obtain a digital label corresponding to each angle value;
forming an observed value sequence of the vehicle motion track based on each digital label;
the sequence generation unit further includes: the association table determining module is specifically used for equally dividing the circumferential direction into 16 parts, and carrying out degree labeling counterclockwise from the positive direction of the abscissa to obtain degree values of each part; dividing the rectangular plane coordinate system into 16 parts equally, and labeling the parts clockwise according to numbers 1-16 in the negative direction of the abscissa to obtain the numerical labels of the parts; associating each degree value with a digital label according to positions to form an angle coding association table;
the probability determination module is used for obtaining corresponding forward probabilities from the observation value sequence through at least one preset driving action model and determining the maximum forward probability from the forward probabilities;
and the action determining module is used for determining the driving action of the obstacle vehicle according to the judgment result of the maximum forward probability.
6. A vehicle, characterized by comprising:
one or more sensors for collecting relative position information of an obstacle vehicle;
one or more controllers;
a storage device for storing one or more programs,
when executed by the one or more controllers, cause the one or more controllers to implement the method of any one of claims 1-4.
7. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-4.
CN201910959765.0A 2019-10-10 2019-10-10 Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium Active CN111284485B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910959765.0A CN111284485B (en) 2019-10-10 2019-10-10 Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910959765.0A CN111284485B (en) 2019-10-10 2019-10-10 Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium

Publications (2)

Publication Number Publication Date
CN111284485A CN111284485A (en) 2020-06-16
CN111284485B true CN111284485B (en) 2021-06-18

Family

ID=71020639

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910959765.0A Active CN111284485B (en) 2019-10-10 2019-10-10 Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium

Country Status (1)

Country Link
CN (1) CN111284485B (en)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111797741A (en) * 2020-06-24 2020-10-20 中国第一汽车股份有限公司 Vehicle detection method, device, vehicle and storage medium
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
CN111982143B (en) * 2020-08-11 2024-01-19 北京汽车研究总院有限公司 Vehicle and vehicle path planning method and device
US11577732B2 (en) * 2020-10-28 2023-02-14 Argo AI, LLC Methods and systems for tracking a mover's lane over time
CN113283647B (en) * 2021-05-19 2023-04-07 广州文远知行科技有限公司 Method and device for predicting obstacle track and automatic driving vehicle
CN113296119B (en) * 2021-05-24 2023-11-28 江苏盛海智能科技有限公司 Unmanned obstacle avoidance driving method and terminal based on laser radar and UWB array
CN113296118B (en) * 2021-05-24 2023-11-24 江苏盛海智能科技有限公司 Unmanned obstacle detouring method and terminal based on laser radar and GPS
CN113734198B (en) * 2021-09-03 2023-04-07 智己汽车科技有限公司 Method and device for acquiring relative course of target
CN115077549B (en) * 2022-06-16 2024-04-26 南昌智能新能源汽车研究院 Vehicle state tracking method, system, computer and readable storage medium
CN114802233B (en) * 2022-06-22 2022-09-16 小米汽车科技有限公司 Vehicle control method, device, electronic device and storage medium
CN115158331A (en) * 2022-07-12 2022-10-11 东风柳州汽车有限公司 Method, device, equipment and storage medium for preventing passengers from dizziness

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8948955B2 (en) * 2010-10-05 2015-02-03 Google Inc. System and method for predicting behaviors of detected objects
CN105530061A (en) * 2016-01-22 2016-04-27 江苏大学 Driving behavior prediction system of forward direction vehicle driver based on fusion of vehicle networking and image
CN105528593A (en) * 2016-01-22 2016-04-27 江苏大学 Forward vehicle driver driving behavior prediction system and prediction method
CN105711586A (en) * 2016-01-22 2016-06-29 江苏大学 Front collision avoiding system based on driving behavior of front vehicle driver and vehicle collision avoiding algorithm
CN106564496A (en) * 2016-10-19 2017-04-19 江苏大学 Reconstruction method for security environment envelope of intelligent vehicle based on driving behaviors of preceding vehicle
CN107697070A (en) * 2017-09-05 2018-02-16 百度在线网络技术（北京）有限公司 Driving behavior Forecasting Methodology and device, unmanned vehicle
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
CN110146100A (en) * 2018-02-13 2019-08-20 华为技术有限公司 Trajectory predictions method, apparatus and storage medium

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2845779B1 (en) * 2013-09-09 2018-08-01 Honda Research Institute Europe GmbH Driving assistance technique for active vehicle control

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8948955B2 (en) * 2010-10-05 2015-02-03 Google Inc. System and method for predicting behaviors of detected objects
CN105530061A (en) * 2016-01-22 2016-04-27 江苏大学 Driving behavior prediction system of forward direction vehicle driver based on fusion of vehicle networking and image
CN105528593A (en) * 2016-01-22 2016-04-27 江苏大学 Forward vehicle driver driving behavior prediction system and prediction method
CN105711586A (en) * 2016-01-22 2016-06-29 江苏大学 Front collision avoiding system based on driving behavior of front vehicle driver and vehicle collision avoiding algorithm
CN106564496A (en) * 2016-10-19 2017-04-19 江苏大学 Reconstruction method for security environment envelope of intelligent vehicle based on driving behaviors of preceding vehicle
CN107697070A (en) * 2017-09-05 2018-02-16 百度在线网络技术（北京）有限公司 Driving behavior Forecasting Methodology and device, unmanned vehicle
CN110146100A (en) * 2018-02-13 2019-08-20 华为技术有限公司 Trajectory predictions method, apparatus and storage medium
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

Also Published As

Publication number Publication date
CN111284485A (en) 2020-06-16

Similar Documents

Publication Publication Date Title
CN111284485B (en) Method and device for predicting driving behavior of obstacle vehicle, vehicle and storage medium
CN113272830B (en) Trajectory representation in behavior prediction system
EP3526737B1 (en) Neural network system for autonomous vehicle control
US20200209857A1 (en) Multimodal control system for self driving vehicle
US11545033B2 (en) Evaluation framework for predicted trajectories in autonomous driving vehicle traffic prediction
US10678248B2 (en) Fast trajectory planning via maneuver pattern selection
JP2020132151A (en) Prediction method for travel locus of vehicle, device, terminal, storage medium and program
US20190084571A1 (en) Driving scenario based lane guidelines for path planning of autonomous driving vehicles
JP7048456B2 (en) Learning devices, learning methods, and programs
CN113267199B (en) Method and device for planning driving track
CN109947090A (en) Non- chocking limit for automatic driving vehicle planning
CN110389582B (en) Tracking objects with multiple cues
US11544556B2 (en) Learning device, simulation system, learning method, and storage medium
CN113655789A (en) Path tracking method, device, vehicle and storage medium
CN115900738A (en) Path planning method, device, vehicle and storage medium
US20210300414A1 (en) Vehicle control method, vehicle control device, and storage medium
US11667281B2 (en) Vehicle control method, vehicle control device, and storage medium
US20210300415A1 (en) Vehicle control method, vehicle control device, and storage medium
CN114537265B (en) Steering lamp control method, device and equipment
US20240199077A1 (en) Motion planning of an autonomous vehicle, motion planning system and vehicle with a motion planning system
US20220326714A1 (en) Unmapped u-turn behavior prediction using machine learning
US20230415766A1 (en) Lane segment clustering using hybrid distance metrics
US20240075923A1 (en) Systems and methods for deweighting veering margins based on crossing time
CN117681908A (en) Track prediction method, track prediction device, vehicle and storage medium
CN115116236A (en) Method and device for predicting driving behavior, storage medium and electronic device

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
GR01 Patent grant
GR01 Patent grant