CN111391848B - Automatic driving vehicle lane changing method - Google Patents
Automatic driving vehicle lane changing method Download PDFInfo
- Publication number
- CN111391848B CN111391848B CN202010137162.5A CN202010137162A CN111391848B CN 111391848 B CN111391848 B CN 111391848B CN 202010137162 A CN202010137162 A CN 202010137162A CN 111391848 B CN111391848 B CN 111391848B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- driving
- degree
- current
- lane change
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 230000008859 change Effects 0.000 claims abstract description 110
- 241000282414 Homo sapiens Species 0.000 claims abstract description 22
- 230000001133 acceleration Effects 0.000 claims description 45
- 238000005065 mining Methods 0.000 abstract description 6
- 230000008901 benefit Effects 0.000 abstract description 3
- 238000009826 distribution Methods 0.000 description 23
- 238000010586 diagram Methods 0.000 description 19
- 230000006399 behavior Effects 0.000 description 12
- 230000008569 process Effects 0.000 description 6
- 230000007613 environmental effect Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000007935 neutral effect Effects 0.000 description 2
- 230000000052 comparative effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
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
- B60W30/18—Propelling the vehicle
- B60W30/18009—Propelling the vehicle related to particular drive situations
- B60W30/18163—Lane change; Overtaking manoeuvres
Landscapes
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
- Traffic Control Systems (AREA)
- Feedback Control In General (AREA)
Abstract
A lane changing method for an automatic driving vehicle belongs to the technical field of automatic driving. The invention aims to establish a psychological factor model capable of reflecting the patience and the courtesy of human beings by mining the traffic big data, and integrate the psychological factor model into a lane change decision of an automatic driving vehicle. The invention constructs a model reflecting the patience degree and a model reflecting the courtesy degree when the automatic driving vehicle follows a slower vehicle ahead, and judges whether to change lanes or not according to the patience threshold value and the courtesy coefficient. The lane change decision method of the automatic driving vehicle, which considers human driving psychological factors, has the advantages of clear structure and easy realization, can be widely applied to various complex traffic environments, and can improve the effectiveness and safety of the lane change decision so as to improve the traffic passing efficiency. Human psychological factors (such as patience, courtesy and the like) are integrated into the lane change decision, and a lane change decision framework which can consider the psychological factors is constructed, so that the usability, the effectiveness and the safety of the lane change decision are improved, and the traffic efficiency of a traffic environment is improved.
Description
Technical Field
The invention belongs to the technical field of automatic driving.
Background
Autonomous vehicles have a rich history, first appearing in the 80's of the 20 th century, and have gained rapid development over the past forty years. The automatic driving vehicle carries an advanced sensing system, an intelligent control system and an accurate execution system, and has the functions of intelligent sensing, autonomous decision making, accurate execution and the like under a certain traffic environment so as to realize safe, comfortable, efficient and energy-saving automatic driving.
The lane change of the vehicle is a basic and important driving behavior, and relates to the driving efficiency and the driving safety of the vehicle. With the improvement of the driving automation degree, the automatic lane changing of the vehicle becomes a necessary function of the automatic driving vehicle. The automatic driving vehicle is in a mixed traffic scene coexisting with human driving vehicles for a long time, so that a driving decision system of the automatic driving vehicle needs to be integrated with factors such as psychology and behaviors of human driving and make relevant decisions similar to human personalized driving, and the traffic efficiency of the vehicle in the mixed traffic environment is further improved.
Patent CN110562258A discloses a vehicle lane change decision method based on a vehicle state space diagram. The method comprises the steps of constructing a vehicle state space diagram based on environmental information and vehicle information; acquiring decision information of a plurality of lane changing actions based on the state space diagram; a lane-change maneuver is then determined based on the decision-making information and the vehicle state information for the plurality of lane-change maneuvers. However, the implementation of the method is based on the deep learning algorithm, the calculation amount is large, the requirement on a vehicle-mounted processor is high, and the method is not easy to be practically applied.
Patent CN110298131A discloses a method for establishing an automatic driving lane change decision model based on a game in a hybrid driving environment. The invention establishes a multi-step dynamic game framework; defining potential conflict points of the lane changing vehicle and the target lane retreat; establishing starting and stopping condition criteria of the game according to the initial information, the strategy and the game steps of the vehicle; and the vehicles use respective strategies and acceleration selection methods to carry out dynamic game until the lane change termination condition is met. However, in the invention, the dynamic information of the vehicle does not take into account the dynamic change and prediction of the acceleration of the vehicle under different decisions, so that when the vehicle makes a decision, the difference between the environment where the vehicle is located and the decision is large when the vehicle performs the decision due to the change of the surrounding environment may not be well handled.
Patent CN109835339A discloses a lane change decision method. The invention obtains the motion information of the vehicle and the environmental vehicle, and divides the driving environment of the vehicle into different areas; determining a target vehicle in each area and the relative distance and the relative speed between the target vehicle and the vehicle; determining lane change zone bits (safe state and unsafe state) of the vehicle in each area; and determining a lane change decision according to the lane change flag bit. However, the motion information acquired by the present invention includes only the relative distance information and the relative speed information, and only expresses the current positional relationship between the host vehicle and the environmental vehicle, and the change in the positional relationship between the host vehicle and the environmental vehicle cannot be predicted well.
At present, the lane change decision of the automatic driving vehicle lacks the comprehensive consideration of the psychological factors of driving, and under the complex mixed traffic environment, the automatic driving vehicle can sometimes make invalid even 'tricky' lane change behavior, so that the traffic efficiency is reduced, and the driving safety is reduced.
Disclosure of Invention
The invention aims to establish a psychological factor model capable of reflecting the patience and the courtesy of human beings by mining the traffic big data, and integrate the psychological factor model into a lane change decision of an automatic driving vehicle.
The method comprises the following steps:
step one, using alphapIndicates the heart-fast threshold, vdesAnd vhRepresenting the desired and actual speed of the current vehicle, Nt1Representing the moment at which the current vehicle starts to follow the slower vehicle ahead, Nt2Representing the time when the current vehicle is expected to begin performing a lane change, a model may be constructed that reflects the degree of patience of the autonomous vehicle when following a slower vehicle ahead:
when inequality (1) of the patience model is established, the present vehicle cannot tolerate the present driving environment;
step two, c represents the current vehicle, o represents the rear following vehicle before the lane change of c, and n represents the rear following vehicle after the lane change of cA rear following vehicle ofc、anAnd aoRespectively represents the acceleration of the vehicles c, n and o before changing lanes c,
and
respectively representing the predicted acceleration values of c, n and o after the lane change of the vehicle c, p representing a gift coefficient, and delta athRepresenting the current vehicle gifts level threshold, a model can be constructed that reflects the gifts level of the autonomous vehicle to other surrounding vehicles when making lane change decisions:
assuming that the acceleration of the current vehicle following the vehicle ahead is constant during the lane change of the vehicle, and the courtesy degree threshold value Delta athSet to 0, thus simplifying the gifts model to:
step three, adopting related vehicle natural driving big data to relate key parameter alpha to the modelpCalibrating p;
setting a sensor system, and acquiring running speed and acceleration information of the automatic driving vehicle and surrounding vehicles;
step five, setting a driving strategy module for the vehicle to longitudinally follow the front vehicle, and generating the acceleration predicted values of the current vehicle and the vehicles in the surrounding environment, namely
And
step six, setting a driving mode selection module of the automatic driving system, wherein the module can beSetting different driving modes according to different driving styles, and updating the patience threshold alpha according to the selected driving modepA yield coefficient p;
step seven, inputting the running speed of the current automatic driving vehicle and the surrounding vehicles into a tolerance degree recognition module to obtain the tolerance degree of the automatic driving system under the current driving environment and driving mode, namely predicting the time for completing the lane change of the vehicle to be T from the current moment, and according to the result that the vehicle can be driven in a lane change mode
Calculating the tolerance degree alpha of the automatic driving system under the current driving environment and the driving modep′；
Step eight, calculating the tolerance degree alpha obtained in the step sevenp' Heart tolerance threshold alpha updated with step sixpComparing, if the current endurance degree is not greater than the endurance threshold value, not changing the channel, otherwise executing the step nine;
step nine, the acceleration of the current automatic driving vehicle and the running acceleration of the surrounding vehicles, the gift coefficient p updated in the step six and the predicted acceleration value obtained in the step five
Input the recognition module of the gift degree according to
Calculating the deviation degree Delta a of the courtesy degree of the automatic driving system under the current driving environment and the driving modeth′；
Step ten, the courtesy degree obtained in the step nine is shifted by a degree delta ath' and the etiquette threshold Deltaa set in step twothAnd comparing the current courtesy degree deviation degree with 0, if the current courtesy degree deviation degree is not larger than the courtesy threshold value of the human driver, indicating that the lane change driving in the current driving environment does not meet the requirement of the current driving mode on the courtesy degree, and making a decision not to change lanes, otherwise making a decision to change lanes.
The lane change decision method of the automatic driving vehicle, which considers human driving psychological factors, has the advantages of clear structure and easy realization, can be widely applied to various complex traffic environments, and can improve the effectiveness and safety of the lane change decision so as to improve the traffic passing efficiency. Human psychological factors (such as patience, courtesy and the like) are integrated into the lane change decision, and a lane change decision framework which can consider the psychological factors is constructed, so that the usability, the effectiveness and the safety of the lane change decision are improved, and the traffic efficiency of a traffic environment is improved.
Drawings
FIG. 1 is a process flow diagram of the present invention;
FIG. 2 is a schematic diagram of the relative position of an autonomous vehicle and surrounding vehicles in accordance with the present invention;
FIG. 3 is a driving patience threshold probability density distribution diagram obtained by mining in Safety Pilot Model Deployment (SPMD) traffic big data in the present invention;
FIG. 4 is a probability density distribution diagram of driving gifts coefficient mined from Safety Pilot Model Deployment (SPMD) traffic big data in the present invention;
FIG. 5 is a graph of vehicle speed, acceleration, and Δ a between a determined autonomous vehicle and a surrounding environmentthUnder the condition of 0, distribution graphs of lane change decisions and lane non-change decisions made by different patience thresholds, different courtesy degrees and the automatic driving vehicle;
FIG. 6 is an exemplary diagram of a lane change decision made using an autonomous vehicle lane change decision module incorporating human psychological factors;
FIG. 7 is an exemplary diagram of a lane change decision-making module for an autonomous vehicle incorporating human psychological factors.
Detailed Description
The invention is described in detail below with reference to the attached drawing figures:
step one, the analysis of the actual driving data and the driving behavior shows that the driving tolerance degree is directly related to the difference between the actual vehicle speed and the expected vehicle speed of the current vehicle, when the difference exceeds a certain threshold value, the current driving environment cannot be tolerated by the vehicle, and the threshold value is described as a tolerance threshold value. At alphapIndicates the heart-fast threshold, vdesAnd vhRepresenting the current vehicleDesired and actual vehicle speeds, Nt1Representing the moment at which the current vehicle starts to follow the slower vehicle ahead, Nt2Representing the time when the current vehicle is expected to begin performing a lane change, a model may be constructed that reflects the degree of patience of the autonomous vehicle when following a slower vehicle ahead:
when inequality (1) of the patience model is established, it indicates that the current vehicle cannot tolerate the current driving environment, and the idea of lane change to lane driving under better driving conditions is generated.
And step two, analyzing the actual driving data and the driving behaviors to know that the courtesy degree of the vehicle is directly related to the negative influence of the vehicle on other surrounding vehicles in the lane changing process, the negative influence is intuitively expressed as the deceleration operation and the deceleration degree of the vehicle driving behind the vehicle caused by the lane changing behavior of the current vehicle before and after the lane changing of the current vehicle, and the smaller the deceleration operation and the smaller the deceleration degree of the rear vehicle are, the higher the courtesy degree of the current vehicle is. Taking FIG. 2 as an example, c represents the current vehicle, o represents the rear following vehicle before the lane change c, n represents the rear following vehicle after the lane change c, ac、anAnd aoRespectively represents the acceleration of the vehicles c, n and o before changing lanes c,
and
respectively representing the predicted values of the acceleration of c, n and o after the lane change of the vehicle c, p representing a courtesy coefficient for reflecting the current courtesy degree of the vehicle, and delta athRepresenting the current vehicle gifts level threshold, a model can be constructed that reflects the gifts level of the autonomous vehicle to other surrounding vehicles when making lane change decisions:
the comity model (2) considers the acceleration state of the running vehicle around the automatic driving vehicle in the lane changing process, and balances the acceleration change value of the vehicle in the lane changing process through the comity coefficient p, so that different comity degrees in the lane changing process are reflected. A courtesy lane change operation is to ensure that the smaller the change of the acceleration of the vehicles is, the better the change is, and further ensure the macroscopic traffic passing efficiency. The p value range of the giving coefficient is (- ∞, + ∞), when the value is negative, the giving coefficient represents that the giving degree of the running vehicle is not high, and is more giving, the smaller the numerical value is, and when the value is positive, the giving coefficient represents that the giving degree of the running vehicle is higher, and is more giving, and the larger the numerical value is, the more giving is. Gifting degree threshold value deltaathThe value range is (— ∞, + ∞), when the threshold value is a negative number, it indicates that the current vehicle can decelerate other surrounding vehicles within a certain degree, that is, the judgment condition on whether the driving behavior of the vehicle is reckless is relatively loose, and the smaller the numerical value is, the more loose the numerical value is, when the threshold value is a positive number, the judgment condition on whether the driving behavior of the current vehicle is courtesy is relatively strict, and the larger the numerical value is, the more strict the numerical value is, and when the threshold value is 0, it indicates that the driving behavior of the current vehicle is reckless, and the courtesy degree is judged to be neutral and unbiased.
Wherein, considering the simplified calibration workload, the acceleration of the current vehicle following the vehicle in front of the vehicle is assumed to be unchanged during the lane change period, and the courtesy degree threshold value Delta athSet to 0, thus simplifying the gifts model to:
thirdly, adopting relevant vehicle natural driving big data to relate a key parameter alpha to the modelpAnd p is calibrated. 10 ten thousand lane change events are extracted from Safety traffic Model Deployment big data (the data is network public data: natural driving data containing 125 vehicles, 4 years and 320 kilometers, https:// category. data. gov/dataset/Safety-Pilot-Model-Deployment-data), and for key parameters related to a patience Model and a courtesy Model,i.e. the endurance threshold alphapAnd a gift factor p. Because the driving performance of the vehicle is influenced by differences of traffic road conditions, regions, culture and the like to a certain extent, the two parameters have certain regional differences. Wherein, aiming at the data of the Michigan area of America, the heart-fast model threshold value alphapThe probability density distribution under big data better approximates the generalized extreme distribution, as shown in FIG. 3; the variation range of the gift coefficient p is [ -2,2]Its probability density distribution under big data better approximates the t location-scale distribution, as shown in FIG. 4.
Setting a sensor system, and acquiring running speed and acceleration information of the automatic driving vehicle and surrounding vehicles; references to sensor systems: (
Jelena,Nenad
and Vujo
"Sensors and sensor fusion in autonomous vehicles.″2018 26th Telecommunications Forum(TELFOR).IEEE,2018.APA)。
Step five, a Vehicle longitudinal following front Vehicle driving strategy module is arranged, a mature strategy (H.H.Yang and H.Peng, Development of an audible car-following driver model, Vehicle System Dynamics, vol.48, No.6, pp.751-773,2010) is built in, and the module is used for generating predicted acceleration values of the current Vehicle and the surrounding Vehicle, namely the predicted acceleration values
And
step six, setting a driving mode selection module of the automatic driving system, wherein the module can set different driving modes according to different driving styles and update the patience threshold value alpha according to the selected driving modepA yield coefficient p;
the driving mode is preset to be a plurality of selectable modes from a reckless mode to a concession mode according to the driving psychological factors, and different modes correspond to different patience thresholds and concession coefficients, as shown in fig. 4.
Step seven, inputting the running speed of the current automatic driving vehicle and the surrounding vehicles into a tolerance degree recognition module, calculating the tolerance degree of the automatic driving system under the current driving environment and driving mode, namely predicting the time for completing the lane change of the vehicle to be T from the current moment, and according to the result that the vehicle can be driven by the automatic driving system in the current driving environment and driving mode
Calculating the tolerance degree alpha of the automatic driving system under the current driving environment and the driving modep′。
Step eight, calculating the tolerance degree alpha obtained in the step sevenp' Heart tolerance threshold alpha updated with step sixpAnd comparing, if the current endurance degree is not greater than the endurance threshold value, not changing the channel, otherwise executing the step nine.
Step nine, the acceleration of the current automatic driving vehicle and the running acceleration of the surrounding vehicles, the gift coefficient p updated in the step six and the predicted acceleration value obtained in the step five
Input the recognition module of the gift degree according to
Calculating the deviation degree Delta a of the courtesy degree of the automatic driving system under the current driving environment and the driving modeth′。
Step ten, the courtesy degree deviation degree delta a obtained by the calculation of the step nineth' and the etiquette threshold Deltaa set in step twothComparing the current courtesy degree deviation degree with 0, if the current courtesy degree deviation degree is not larger than a courtesy threshold value of a human driver, indicating that lane change driving under the current driving environment does not meet the requirement of the current driving mode on the courtesy degree, and making a decision not to change lanes, otherwise making a decision to change lanes;
FIG. 1 is a flow chart of a lane change decision method of an automatic driving vehicle incorporating human psychological factors of the present invention, wherein a sensor system installed in the automatic driving vehicle collects driving information (speed, acceleration) of a current vehicle and surrounding vehicles and transmits the information to a lane change decision module of the automatic driving vehicle incorporating human psychological factors.
2. The running information of the automatic driving vehicle and the surrounding environment vehicle acquired by the sensor system in the step 1 is input into a longitudinal following strategy module to obtain the predicted value of the acceleration of the automatic driving vehicle after the automatic driving vehicle is expected to complete lane change
Predicted acceleration value of vehicle relative to surrounding environment
3. The automatic driving vehicle updates the patience threshold alpha in the current mode according to the driving mode selected by the vehicle occupantpA yield coefficient p and a yield threshold Δ ath。
4. Firstly inputting the speed information of the automatic driving vehicle and the surrounding environment vehicle acquired in the step 1 into a tolerance degree identification module, calculating the tolerance degree under the current driving environment and driving mode and comparing the calculated tolerance degree with the updated tolerance threshold value alpha in the step 3pComparing, if the calculated tolerance degree is larger than alphapIf the vehicle speed information is the same as the predicted acceleration value in the step 2, the vehicle speed information is obtained in the step 1, and the predicted acceleration value is obtained in the step 2
And
and inputting a courtesy degree identification module, continuously judging the feasibility of lane change driving, and if not, making a decision not to change lanes, wherein the next lane change decision (5) is visible.
5. CourierAfter the degree identification module receives the speed information of the automatic driving vehicle and the surrounding vehicle acquired in the step 1 and the predicted acceleration value predicted in the step 2, the gifting coefficient p under the current driving mode updated in the step 3 is synthesized to calculate the deviation degree of the gifting degree under the current driving environment and the driving mode, and the deviation degree and the gifting threshold value delta a updated in the step 3 are comparedthComparing if the calculated courtesy degree is larger than delta athIf the lane change decision is not made, the lane change decision can be made if the influence of the automatic driving vehicle lane change driving on the surrounding environment vehicle meets the requirement of the courtesy degree under the current driving environment and driving mode, namely the lane change driving is feasible.
FIG. 3 is a probability density distribution diagram of driving tolerance threshold value obtained by mining in Safety Pilot Model Deployment (SPMD) traffic big data according to the present invention
1. In this figure, the horizontal axis represents the tolerance level calculated in the tolerance model (1), the vertical axis represents the probability density of the tolerance level in the data, and the higher the tolerance level is, the lower the probability density is, i.e., the less the data is;
2. in the graph, a histogram is vehicle driving data in the large SPMD traffic data, ". o" represents an approximate trend of the data in a generalized pareto distribution, "+" represents an approximate trend of the data in an exponential distribution, ". -" represents an approximate trend of the data in a generalized extreme distribution;
3. in the graph, the vehicle driving data in the region better approximates the generalized extreme value distribution as seen from the comparison of the histogram with the approximate trend curves of three different distributions.
FIG. 4 is a probability density distribution diagram of driving benefit coefficient obtained by mining in Safety Pilot Model Deployment (SPMD) traffic big data according to the present invention
1. In the figure, the horizontal axis represents the comity threshold Δ athWhen the result is 0, the result calculated by the result model (2) and the vertical axis is the probability density of the result in the data, the data shows obvious 'middle height and both ends low', that is, the driving behaviors with reckless degree and result degree lower are most in number, and the extreme driving behaviors with reckless degree and result degree higher are few.
2. In the figure, the local gift factor is at the gift threshold Δ athThe value range under the condition of 0 is [ -2,2]；
3. In the figure, the more leftward the value of the giving coefficient is shifted, the smaller the corresponding reckless degree is, and the lower the giving degree is, and conversely, the more rightward the value of the giving coefficient is shifted, the lower the corresponding reckless degree is, and the higher the giving degree is;
4. in the figure, a histogram is vehicle driving data in the large data of the SPMD traffic, wherein ". smallcircle" represents an approximate trend of the data under a tlocation-scale distribution, "…" represents an approximate trend of the data under a logistic distribution, "- -" represents an approximate trend of the data under a normal distribution, and "-" represents an approximate trend of the data under a broad extreme value distribution;
5. in the graph, as can be seen from comparison of the histogram with four approximate trend curves under different distributions, the vehicle driving data gift coefficient in the area is at the gift threshold Δ athThe t-location-scale distribution is well approximated under 0 conditions.
The first practical effect diagram is as follows: FIG. 5 is a graph of vehicle speed, acceleration, and Δ a between a determined autonomous vehicle and a surrounding environmentthDistribution diagram of lane change decision and lane non-change decision made by different patience thresholds, different courts and automatic driving vehicles under the condition of 0
1. In this example, the positional relationship between the autonomous vehicle and the surrounding vehicle will be described later with reference to fig. 2;
2. in this example, the autonomous vehicle c and the required surrounding vehicle travel information and prediction information are as follows:
1) the current running speed of the automatic driving vehicle c is vh＝10m/s；
2) The desired speed v of the autonomous vehicle cdes＝25m/s；
3) The current acceleration of the autonomous vehicle c is ac＝0m/s2；
4) The current acceleration of the rear vehicle after the lane change of the automatic driving vehicle is an＝0.5m/s2；
5) Automatic driving vehicleThe predicted acceleration value of the vehicle after the expected lane change is
6) The predicted value of the acceleration of the rear vehicle after the lane change of the automatic driving vehicle after the expected lane change of the automatic driving vehicle is
3. Substituting the conditions 1) and 2) in the step 2 into the patience model (1) to obtain a smaller patience threshold value alphapThe intention of changing the driving lane of the automatic driving vehicle is easier to generate, and conversely, the greater tolerance threshold value alphapThe automatic driving vehicle is easier to keep running in the current lane state, and the intention of changing the running lane is not easy to generate;
4. the conditions 3), 4), 5) and 6) in the step 2 are brought into the simplified courtesy model (3), so that a driving lane changing decision can be made more easily by a smaller courtesy coefficient p, the automatic driving vehicle can be kept in a current lane state more easily by a larger courtesy coefficient p, the influence on surrounding environment vehicles is reduced as much as possible, and the driving lane changing decision is not made easily;
5. the analysis results of 3 and 4 are combined, and the patient can be treated as the patience threshold value alphapWhen the vehicle is bigger, the driving endurance of the automatic driving vehicle is stronger, and more lane changing-free decisions are made; when the gift coefficient p is larger, the higher the gift degree of the driving of the automatic driving vehicle, the more lane change-free decisions are made.
The practical effect graph II: fig. 6 and 7 are diagrams of comparative experimental examples of making a lane change decision and a non-lane change decision by using an automatic driving vehicle lane change decision module integrated with human psychological factors, respectively, where the experimental conditions are set as follows: desired vehicle speed v of autonomous vehicledes＝18m/s2(about 65 km/h); the tolerance threshold is alpha p500; the gift coefficient adopts the most typical example p which accounts for the highest proportion of the gift coefficient probability density distribution curve and is 0; etiquette threshold value Deltaa th0 represents the neutral, unbiased judgment of the reckless, beneft degree of the driving behavior of the autonomous vehicle
FIG. 6 is an exemplary diagram of a lane change decision made by an autonomous vehicle lane change decision module incorporating human psychological factors
1. The figure has three sub-figures, namely a 1) vehicle speed change curve graph of the automatic driving vehicle and a vehicle in front of the automatic driving vehicle, a 2) distance change curve graph between the automatic driving vehicle and the vehicle in front and a 3) lane mark change curve graph of the automatic driving vehicle;
2. fig. 3) is a graph showing the change of lane marks of the automatically driven vehicle, in the experiment, the lane where the vehicle is located is represented by the lane mark where the center position of the vehicle is located, and when the center position of the vehicle is close to the separation line of the two lanes and moves from one side to the other side, the change of the lane marks of the vehicle is generated, so that the change of the curve is approximate to a step shape;
3. from sub-diagram 3) it can be seen from the lane-marking change curve graph that the autonomous vehicle is driving, that lane change is performed during driving;
3. three subgraphs in the graph are respectively marked with ellipses 1,2 and 3 to form a section of curve before and after lane changing of the automatic driving vehicle;
4. in sub-diagram 1) vehicle speed variation graph of the autonomous vehicle and the vehicle ahead of the autonomous vehicle: the running speed of the automatic driving vehicle is about 13m/s2Far below the expected vehicle speed, the use of the patience model (1) has the intention of automatically driving the vehicle to change the driving lane;
5. in sub-diagram 1) vehicle speed variation graph of the autonomous vehicle and the vehicle ahead of the autonomous vehicle: it can be seen from the area marked by the ellipse 1 that the speed of the automatically driven vehicle is basically unchanged before lane changing, so the acceleration is approximately 0 before lane changing, the vehicle speed is rapidly increased after lane changing, and the acceleration is very large, so the increment of the acceleration of the automatically driven vehicle before and after lane changing is very large, and the calculated courtesy degree of the simplified courtesy model (3) is larger than a courtesy threshold value;
6. and based on the result of 5, under the current driving environment and driving mode, making a lane change decision of the automatic driving vehicle.
FIG. 7 is a diagram of an example of making a lane change-free decision using an autonomous vehicle lane change decision module incorporating human psychological factors
1. The figure has three sub-figures, namely a 1) vehicle speed change curve graph of the automatic driving vehicle and a vehicle in front of the automatic driving vehicle, a 2) distance change curve graph between the automatic driving vehicle and the vehicle in front and a 3) lane mark change curve graph of the automatic driving vehicle;
2. from fig. 3) it can be seen from the lane sign change graph that the autonomous vehicle is traveling, that lane change is not performed in the traveling process of the autonomous vehicle;
3. the method is characterized in that 1) a speed change curve graph of an automatic driving vehicle and a vehicle in front of the automatic driving vehicle is integrated, 2) a distance change curve graph between the automatic driving vehicle and the vehicle in front of the automatic driving vehicle is analyzed, the driving speed of the automatic driving vehicle is almost close to the expected speed, and the automatic driving vehicle does not generate the intention of changing a driving lane by applying a patience model (1);
4. and (4) integrating the results of the step (3), and making a decision of not changing the lane by the automatic driving vehicle under the current driving environment and driving mode.
The invention integrates human psychological factors, so that lane change decision is more consistent with the characteristics of real human driving style, and the traffic efficiency of hybrid traffic can be improved. The method has clear logic and strong generalization capability, is convenient for data acquisition, and can be expanded and applied to hybrid traffic microscopic modeling and energy efficiency and safety evaluation of running vehicles in hybrid traffic environments.
The invention is characterized in that:
1. a patience and courtesy model reflecting human driving psychological factors is established, the model is simple in form and easy to realize, and can be well integrated into the design of automatic driving lane change decision;
2. by mining the natural driving big data, the random probability density distribution of the patience and courtesy degree is counted, and key model parameters reflecting different driving styles are calibrated;
3. an automatic driving lane change decision-making framework which is integrated with psychological factors is constructed, and the method has strong universality.
Claims (1)
1. A lane changing method for an automatic driving vehicle is characterized by comprising the following steps: the method comprises the following steps:
step one, using alphapIndicates the heart-fast threshold, vdesAnd vhRepresenting the desired and actual speed of the current vehicle, Nt1Representing the moment at which the current vehicle starts to follow the slower vehicle ahead, Nt2Representing the time when the current vehicle is expected to begin performing a lane change, a model may be constructed that reflects the degree of patience of the autonomous vehicle when following a slower vehicle ahead:
when inequality (1) of the patience model is established, the present vehicle cannot tolerate the present driving environment;
step two, c represents the current vehicle, o represents the rear following vehicle before the lane change of c, n represents the rear following vehicle after the lane change of c, ac、anAnd aoRespectively represents the acceleration of the vehicles c, n and o before changing lanes c,
and
respectively representing the predicted acceleration values of c, n and o after the lane change of the vehicle c, p representing a gift coefficient, and delta athRepresenting the current vehicle gifts level threshold, a model can be constructed that reflects the gifts level of the autonomous vehicle to other surrounding vehicles when making lane change decisions:
assuming that the acceleration of the current vehicle following the vehicle ahead is constant during the lane change of the vehicle, and the courtesy degree threshold value Delta athSet to 0, thus simplifying the gifts model to:
step three, adopting related vehicle natural driving big data to relate key parameter alpha to the modelpCalibrating p;
setting a sensor system, and acquiring running speed and acceleration information of the automatic driving vehicle and surrounding vehicles;
step five, setting a driving strategy module for the vehicle to longitudinally follow the front vehicle, and generating the acceleration predicted values of the current vehicle and the vehicles in the surrounding environment, namely
And
step six, setting a driving mode selection module of the automatic driving system, wherein the module can set different driving modes according to different driving styles and update the patience threshold value alpha according to the selected driving modepA yield coefficient p;
step seven, inputting the running speed of the current automatic driving vehicle and the surrounding vehicles into a tolerance degree recognition module to obtain the tolerance degree of the automatic driving system under the current driving environment and driving mode, namely predicting the time for completing the lane change of the vehicle to be T from the current moment, and according to the result that the vehicle can be driven in a lane change mode
Calculating the tolerance degree alpha of the automatic driving system under the current driving environment and the driving modep′；
Step eight, calculating the tolerance degree alpha obtained in the step sevenp' Heart tolerance threshold alpha updated with step sixpComparing, if the current endurance degree is not greater than the endurance threshold value, not changing the channel, otherwise executing the step nine;
step nine, the acceleration of the current automatic driving vehicle and the running acceleration of the surrounding vehicles, the gift coefficient p updated in the step six and the predicted acceleration value obtained in the step five
Input the recognition module of the gift degree according to
Calculating the deviation degree Delta a of the courtesy degree of the automatic driving system under the current driving environment and the driving modeth′；
Step ten, the courtesy degree obtained in the step nine is shifted by a degree delta ath' and the etiquette threshold Deltaa set in step twothAnd comparing the current courtesy degree deviation degree with 0, if the current courtesy degree deviation degree is not larger than the courtesy threshold value of the human driver, indicating that the lane change driving in the current driving environment does not meet the requirement of the current driving mode on the courtesy degree, and making a decision not to change lanes, otherwise making a decision to change lanes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010137162.5A CN111391848B (en) | 2020-03-02 | 2020-03-02 | Automatic driving vehicle lane changing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010137162.5A CN111391848B (en) | 2020-03-02 | 2020-03-02 | Automatic driving vehicle lane changing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111391848A CN111391848A (en) | 2020-07-10 |
| CN111391848B true CN111391848B (en) | 2022-03-08 |
Family
ID=71419717
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010137162.5A Active CN111391848B (en) | 2020-03-02 | 2020-03-02 | Automatic driving vehicle lane changing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111391848B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114506324B (en) * | 2020-10-23 | 2024-03-15 | 上海汽车集团股份有限公司 | Lane decision method and related device |
| CN113495563B (en) * | 2021-06-10 | 2022-09-20 | 吉林大学 | Traffic vehicle lane change decision planning method for automatic driving virtual test |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104260725A (en) * | 2014-09-23 | 2015-01-07 | 北京理工大学 | Intelligent driving system with driver model |
| CN105460015A (en) * | 2014-09-24 | 2016-04-06 | 株式会社万都 | Automatic driving controlling system and method |
| CN108595823A (en) * | 2018-04-20 | 2018-09-28 | 大连理工大学 | A kind of computational methods of Autonomous Vehicles lane-change strategy that combining driving style and theory of games |
| CN108688657A (en) * | 2017-03-31 | 2018-10-23 | 福特全球技术公司 | It virtually can turning path |
| CN110077403A (en) * | 2018-01-26 | 2019-08-02 | 丰田自动车株式会社 | Controller of vehicle |
| CN110298131A (en) * | 2019-07-05 | 2019-10-01 | 西南交通大学 | Automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment |
| CN110341712A (en) * | 2019-06-11 | 2019-10-18 | 汉腾汽车有限公司 | A kind of automatic Pilot system for recognizing lane-changing intention and its method |
| CN110356404A (en) * | 2019-05-28 | 2019-10-22 | 吉林大学 | A kind of intelligent driving system for having the function of autonomous lane-change and improving laterally security |
| CN110667585A (en) * | 2018-06-15 | 2020-01-10 | 本田技研工业株式会社 | Vehicle control device, vehicle control method, and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20190054374A (en) * | 2017-11-13 | 2019-05-22 | 한국전자통신연구원 | Autonomous drive learning apparatus and method using drive experience information |
-
2020
- 2020-03-02 CN CN202010137162.5A patent/CN111391848B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104260725A (en) * | 2014-09-23 | 2015-01-07 | 北京理工大学 | Intelligent driving system with driver model |
| CN105460015A (en) * | 2014-09-24 | 2016-04-06 | 株式会社万都 | Automatic driving controlling system and method |
| CN108688657A (en) * | 2017-03-31 | 2018-10-23 | 福特全球技术公司 | It virtually can turning path |
| CN110077403A (en) * | 2018-01-26 | 2019-08-02 | 丰田自动车株式会社 | Controller of vehicle |
| CN108595823A (en) * | 2018-04-20 | 2018-09-28 | 大连理工大学 | A kind of computational methods of Autonomous Vehicles lane-change strategy that combining driving style and theory of games |
| CN110667585A (en) * | 2018-06-15 | 2020-01-10 | 本田技研工业株式会社 | Vehicle control device, vehicle control method, and storage medium |
| CN110356404A (en) * | 2019-05-28 | 2019-10-22 | 吉林大学 | A kind of intelligent driving system for having the function of autonomous lane-change and improving laterally security |
| CN110341712A (en) * | 2019-06-11 | 2019-10-18 | 汉腾汽车有限公司 | A kind of automatic Pilot system for recognizing lane-changing intention and its method |
| CN110298131A (en) * | 2019-07-05 | 2019-10-01 | 西南交通大学 | Automatic Pilot lane-change decision model method for building up under a kind of mixing driving environment |
Non-Patent Citations (6)
| Title |
|---|
| Lu-Zheng Bi ; Jun-Xing Shang ; Guo-Dong Gan.Modeling driver lane changing control with the queuing network-model human processor.《 IEEE/IEE Electronic Library》.2012, * |
| 基于UCWin/Road仿真的高速公路出口预告标志前置距离研究;潘兵宏等;《公路交通科技》;20170815(第08期);全文 * |
| 基于决策树模型的驾驶员期望车速;王晓原等;《计算机应用》;20091231;全文 * |
| 智能车辆高速公路自动变道轨迹规划与控制研究;刘洋;《中国优秀硕士学位论文全文数据库电子期刊工程科技Ⅱ辑》;20191115(第11期);全文 * |
| 网联环境下自主车的换道决策研究;梁文斌;《中国优秀硕士学位论文全文数据库电子期刊工程科技Ⅱ辑》;20190215(第2期);全文 * |
| 车道变换决策仿真模型;王晓原等;《计算机工程与应用》;20100601(第16期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111391848A (en) | 2020-07-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Driving style analysis using primitive driving patterns with Bayesian nonparametric approaches | |
| CN109598934B (en) | Rule and learning model-based method for enabling unmanned vehicle to drive away from high speed | |
| CN111775949B (en) | Personalized driver steering behavior auxiliary method of man-machine co-driving control system | |
| US6873911B2 (en) | Method and system for vehicle operator assistance improvement | |
| CN113291308B (en) | Vehicle self-learning lane-changing decision-making system and method considering driving behavior characteristics | |
| CN112347567A (en) | Vehicle intention and track prediction method | |
| CN112840350A (en) | Autonomous vehicle planning and prediction | |
| CN110103956A (en) | Automatic overtaking track planning method for unmanned vehicle | |
| CN111391848B (en) | Automatic driving vehicle lane changing method | |
| CN112249008B (en) | Unmanned automobile early warning method aiming at complex dynamic environment | |
| CN110568841A (en) | Automatic driving decision method and system | |
| CN111275249A (en) | Driving behavior optimization method based on DQN neural network and high-precision positioning | |
| CN112793576B (en) | Lane change decision method and system based on rule and machine learning fusion | |
| US20220306113A1 (en) | Customization of autonomous-driving lane changes of motor vehicles based on drivers' driving behaviours | |
| CN113255998B (en) | Expressway unmanned vehicle formation method based on multi-agent reinforcement learning | |
| CN113722835B (en) | Personification random lane change driving behavior modeling method | |
| CN114572244A (en) | Intelligent driving and ADAS front vehicle driving behavior recognition and prediction method and module | |
| Wang et al. | Haptic assistive control with learning-based driver intent recognition for semi-autonomous vehicles | |
| CN111775938A (en) | Lane change path planning method, device and system | |
| Guo et al. | Toward human-like behavior generation in urban environment based on Markov decision process with hybrid potential maps | |
| CN115923833A (en) | Personifying decision control method and device for vehicle, vehicle and storage medium | |
| CN114802306A (en) | Intelligent vehicle integrated decision-making system based on man-machine co-driving concept | |
| CN117585017A (en) | Automatic driving vehicle lane change decision method, device, equipment and storage medium | |
| CN107045639B (en) | Following behavior generation method based on time automaton | |
| Griesbach et al. | Prediction of lane change by echo state networks |
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 |