CN116663403A - Signal intersection traffic simulation method based on Bayesian network and following model - Google Patents
Signal intersection traffic simulation method based on Bayesian network and following model Download PDFInfo
- Publication number
- CN116663403A CN116663403A CN202310592097.9A CN202310592097A CN116663403A CN 116663403 A CN116663403 A CN 116663403A CN 202310592097 A CN202310592097 A CN 202310592097A CN 116663403 A CN116663403 A CN 116663403A
- Authority
- CN
- China
- Prior art keywords
- model
- vehicle
- intersection
- driver
- following
- 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.)
- Granted
Links
- 238000004088 simulation Methods 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 title claims abstract description 25
- 230000035945 sensitivity Effects 0.000 claims description 18
- 230000006399 behavior Effects 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000002776 aggregation Effects 0.000 description 3
- 238000004220 aggregation Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/29—Graphical models, e.g. Bayesian networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
-
- G06Q50/40—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Abstract
The invention discloses a signal intersection simulation method based on a Bayesian network and a following model, and belongs to the technical field of traffic simulation and control. Firstly, dividing four typical traffic scenes according to driving behavior characteristics of a driver in different signal lamp states; secondly, four following models are respectively established according to following behavior characteristics under different traffic scenes; then, taking the speed of the target vehicle, the signal lamp status lamp and the distance from the target vehicle to the parking line as influencing factors influencing the decision of the driver to construct a Bayesian network model; furthermore, training a Bayesian network model based on actual vehicle track data and calculating the probability that the driver passes through the intersection without stopping; and finally, switching the following model in real time and dynamically according to the calculated probability. The method integrates the advantages of the probability model and the dynamics model, can better reflect the following behavior of the driver in the traffic scene of the actual signalized intersection, and has better expansibility and robustness.
Description
Technical Field
The invention relates to a vehicle following simulation method in the technical field of traffic simulation and control, in particular to a signal intersection traffic simulation method based on a Bayesian network and a following model.
Background
As the amount of maintenance of automobiles increases year by year, urban traffic congestion becomes increasingly problematic. The signal intersection is used as an important node of the urban road network, and is important for urban traffic operation. However, intersections are also a high-rise area of urban congestion and a difficulty in traffic control. The traffic phenomenon and traffic characteristics near the signalized intersections are very complex due to the influence of the signal lamps and the collection of the multi-directional traffic flows. The following model is considered as an important component of microscopic traffic as an effective means of studying traffic flow and understanding traffic phenomena.
Masashi Sasaki et al studied the impact of single lane signal lamps on traffic flow based on an optimal speed model and traffic flow characteristics under different signal control strategies. Tang Tieqiao and the like propose a traffic flow model considering the influence of signal lamps based on a full speed difference model, and the result proves that the traffic flow model can better reproduce the aggregation and dissipation characteristics of traffic flow. Chen et al have established a vehicle following model describing four phases of movement of the vehicle, taking into account the movement characteristics of the signalized intersection vehicle. In shao wei, li Xiuhai, etc., a signalized intersection vehicle aggregation model was established taking into account the aggregation characteristics of the intersection vehicles. Tang and Yu et al incorporate the green time driver's following characteristics into the following model. Zhang et al analyze the following behavior of the driver in different signal states and build a staged following model based on different signal states. Zhu et al studied the braking process of the signalized intersection vehicle.
In the research of the existing signalized intersection following simulation, most of the following models are used for modeling driving behaviors. Wherein when multiple heel models are involved, the switching between models is by deterministic calculation rules. However, in an actual traffic environment, the decision behavior of the driver has a high degree of heterogeneity and randomness when faced with different following and signal conditions. Therefore, the actual driving behavior and traffic conditions are not well reflected by using only the following model and deterministic rules as the microscopic simulation model of the signalized intersection.
Disclosure of Invention
The invention aims to realize microscopic vehicle simulation which is more in line with the actual signalized intersection. The method of data driving and model driving is adopted to construct the following simulation method which is more in line with the actual signal intersection, specifically, a Bayesian network is used as a decision model of a driver, and 4 following models in different scenes are used as following behavior models of a vehicle.
A signal intersection simulation method based on a Bayesian network and a following model comprises the following steps:
step one: traffic scenarios that vehicles may face in passing through a signalized intersection are analyzed, generalized, and described. All possible traffic scenarios can be divided into the following four scenarios:
scene one: before the stop line of the intersection, the target vehicle is not the head vehicle, the current signal lamp is the red light or the green light end, and the current vehicle cannot pass through the intersection;
scene II: before the stop line of the intersection, the target vehicle is the head vehicle, the current signal lamp is the red light or the green light end, and the current vehicle cannot pass through the intersection;
scene III: before the intersection stops the line, the target vehicle is not the head vehicle, the current signal lamp is at the end of the green light or the red light, and the current vehicle can pass through the intersection without stopping;
scene four: before the intersection stops, the target vehicle is a head vehicle, the current signal lamp is at the end of a green light or a red light, and the current vehicle can pass through the intersection without stopping;
step two: and establishing a vehicle following model under different scenes of the signalized intersection.
By analyzing the behavior characteristics of drivers in different traffic scenes, the following behaviors of the drivers in different scenes are described by using 4 following models respectively.
Scene one: when the target vehicle is not a head vehicle and cannot pass through the intersection without stopping, the following model of the vehicle is as follows:
Ψ(Δv n (t),l n (t))＝min{Δx n (t),l n (t)}
wherein v is n (t) is the speed of the nth vehicle at time t, Δx n (t) is the head space between the nth vehicle and the preceding vehicle (the (n-1) th vehicle), l n (t) is the distance from the nth vehicle to the parking line at time t; v (V) op,scen1 Is the optimal speed of the vehicle in the scene. Alpha 1 ，β 1 Respectively an optimal speed sensitivity coefficient and a speed difference sensitivity coefficient of a driver in the model I, wherein h is a safety distance, v max Is the maximum vehicle speed.
Scene II: when the target vehicle is a head vehicle and cannot pass through the intersection without stopping, the following model of the vehicle is as follows:
wherein alpha is 2 ，β 2 The optimal speed sensitivity coefficient and the speed difference sensitivity coefficient of the driver in the second model are respectively; scene III: when the target vehicle is not a head vehicle and can pass through the intersection without stopping, the following model of the vehicle is as follows:
wherein alpha is 3 ，β 3 The optimal speed sensitivity coefficient and the speed difference sensitivity coefficient of the driver in the model III are respectively;
scene four: when the target vehicle is a head vehicle and can pass through the intersection without stopping, the following model of the vehicle is:
wherein v is exp For maximum desired speed, M is any number large enough that V op (M)＝v exp 。
Step three: establishing a decision model of a driver at the signalized intersection based on the Bayesian network.
Based on actual traffic scene analysis, a Bayesian network model is constructed by taking the speed of the target vehicle, the signal lamp status lamp and the distance from the target vehicle to the parking line as influencing factors influencing the decision of a driver. The joint probability of the Bayesian network is shown as follows:
p[v n (t),l n (t),S(t),J n (t+τ)]
＝p(v n (t)|l n (t),S(t)).p(L n (t)).p(S(t)).p(J n (t+τ)|v n (t),l n (t),S(t))
where S (t) represents the status category of the time signal. J (J) n And (t+tau) is the decision-making behavior of the driver at the moment t+tau, and is the speed reduction to stop and the passing of no stop. In addition, for the convenience of calculation, v n (t)，l n (t), the value of S (t) is also discretized into several categories. In the studies herein S (t), J n (t+τ)、v n (t)、l n (t) is divided into 9, 2, 7 and 9 classes respectively. The corresponding classification interval is as follows:
v n (t)∈[0,2),[2,4),[4,6),[6,8),[8,10),[10,12),[12,v max ) Wherein the unit is m/s.
The corresponding categories are: { Vr1, vr2, vr3, vr4, vr5, vr6, vr7}
l n (t) ∈ [0,10 ], [10,20 ], [20,30 ], [30,45 ], [45,60 ], [60,75 ], [75,90 ], [90,105 ], [105, ++ infinity), wherein the unit is m. The corresponding class value is: { Dr1, dr2, dr3, dr4, dr5, dr6, dr7, dr8, dr9}
S(t)∈[t g,end ,t y,end )[t r,end ,t g,end -20s)[t r,end -20s,t r,end -10s)[t r,end -10s,t r,end -3s)[t r,end -3s,t r,end )[t y,end ,t r,end -20s)[t g,end -20s,t g,end -10s)[t g,end -10s,t g,end -3)[t g,end -3s,t g,end )
Wherein t is r,end ，t g,end ，t y,end The end time of the red light, the green light and the yellow light in a signal period are respectively. The unit is s. The categories of the pair are as follows: { L1, L2, L3, L4, L5, L6, L7, L8, L9}.
J n (t+τ) ∈ {0,1}, 1 is taken when the probability of passing through the vehicle without stopping is greater than 0.5, 0 is taken when it is less than 0.5, the corresponding category is expressed as: { S0, S1}. It should be noted that the classification and the division of the sections can be adjusted according to the specific situation.
Step four: a Bayesian network model is trained based on actual intersection scene vehicle trajectory data.
Step five: the probability p that the driver passes through the intersection without stopping is calculated based on the bayesian network model.
The bayesian network dynamically calculates the probability of whether the driver can pass through the intersection (i.e. the probability of slowing down and stopping or not passing through the signalized intersection) in real time under the current driving state and the signal lamp state. The probability p can be calculated by:
wherein i, j, k represent v n (t)，l n (t), the values of S (t) respectively belong to the corresponding ith, j and k categories. The bayesian network dynamically calculates the probability of whether the driver can pass through the intersection (i.e. the probability of slowing down and stopping or not passing through the signalized intersection) in real time under the current driving state and the signal lamp state.
Step six: when other vehicles exist between the target vehicle and the parking line and p is smaller than 0.5, selecting the first model as a following model of the driver at the current moment; when no other vehicle exists between the target vehicle and the parking line and p is less than 0.5, selecting the second model as a following model of the driver at the current moment; when other vehicles exist between the target vehicle and the parking line and p is more than 0.5, selecting a model III as a following model of the driver at the current moment; when no other vehicle exists between the target vehicle and the parking line and p is more than 0.5, selecting a model IV as a following model of the driver at the current moment;
step seven: and (3) repeating the step five and the step six at each moment during simulation.
Compared with the prior art, the invention has the beneficial effects that:
the invention adopts a data driving and model driving method to construct a signal intersection following simulation method, firstly, the actual traffic condition is simplified and generalized into 4 traffic scenes, and 4 following models are used as the following behavior models of vehicles in each scene. And fully considering the heterogeneity and randomness of the driver, establishing a decision model of the driver based on a Bayesian network, and dynamically calculating the switching probability of the following model in real time. The method integrates the advantages of the probability model and the dynamics model, can better reflect the following behavior of the driver in the traffic scene of the actual signalized intersection, and has better expansibility and robustness.
Description of the drawings:
fig. 1 is a diagram illustrating 4 scenes of vehicle following in a bayesian network and following model-based signal intersection simulation method according to the present invention.
Fig. 2 is a diagram of the bayesian network in the bayesian network and following model-based signal intersection simulation method according to the present invention.
Fig. 3 is a schematic diagram of the result of training a bayesian network using Netica software in the bayesian network and following model based signal intersection simulation method according to the present invention.
Fig. 4 is a specific flowchart of vehicle following model switching based on a bayesian network decision model in the bayesian network and following model-based signal intersection simulation method according to the present invention.
Detailed Description
Specific embodiments of the present disclosure are described in detail below with reference to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating and illustrating the disclosure, are not intended to limit the disclosure.
Step one: analyzing, summarizing and describing traffic scenes that vehicles may face in the process of passing through a signalized intersection;
referring to fig. 1, a generalization of vehicles through intersections into 4 scenarios: in scenario 1, there are other vehicles in front of the target vehicle and they cannot pass through the intersection within the current signal period; in scenario 2, there is no other vehicle in front of the target vehicle, i.e., the target vehicle is the head car, and it cannot pass through the intersection in the current signal period; in scenario 3, there are other vehicles in front of the target vehicle, and they may not stop through the intersection during the current period; in scenario 4, there are no other vehicles in front of the target vehicle, i.e., the target vehicle is the head car, and it may pass through the intersection during the current signal period. In the simulation process, the scene of the vehicle at each moment can be judged in real time.
Step two: setting up a following model of a driver in different scenes of a signalized intersection;
and 4 different heel models are respectively built according to different heel scenes in the step one. For scenario 1, the built heel model is as follows:
wherein the optimal velocity function V op The calculation is performed by the following formula:
wherein H (·) is a sea plug function, and ψ (·) is calculated by the following formula:
Ψ(Δv n (t),l n (t))＝min{Δx n (t),l n (t)}
wherein v is n (t) is the speed of the nth vehicle at time t, Δx n (t) is the head space between the nth vehicle and the preceding vehicle (the (n-1) th vehicle), l n (t) is the distance from the nth vehicle to the parking line at time t; v (V) op And H (·) is a sea plug function and is an optimal speed function of the vehicle under the scene. Alpha 1 ，β 1 Respectively an optimal speed sensitivity coefficient and a speed difference sensitivity coefficient of a driver in the model I, wherein h is a safety distance, v max Is the maximum vehicle speed.
For scenario 2, the following model used in the simulation is set up as follows:
the optimal speed function of the vehicle is calculated by:
wherein alpha is 2 ，β 2 The optimal speed sensitivity coefficient and the speed difference sensitivity coefficient of the driver in the scene 2 are respectively;
for scenario 3, the following model used in the simulation is set up as follows:
the optimal speed function of the vehicle is calculated by:
wherein alpha is 3 ，β 3 The optimal speed sensitivity coefficient and the speed difference sensitivity coefficient of the driver in the model III are respectively;
for scene 4, the following model used in the simulation is set up as follows:
the optimal speed of the vehicle is calculated by:
wherein v is exp For maximum desired speed, M is any number large enough that V op (M)＝v exp 。
Step three: establishing a decision model of a driver at the signalized intersection based on the Bayesian network.
Referring to fig. 2, a target vehicle speed, a target vehicle to stop line distance, a signal lamp status, and a driver decision are selected as node variables in a bayesian network, and a bayesian network-based signalized intersection driver decision model is constructed. The bayesian network model is represented by a joint probability density function of each variable as shown in the following formula:
p[v n (t),l n (t),S(t),J n (t+τ)]
＝p(v n (t)|l n (t),S(t)).p(L n (t)).p(S(t)).p(J n (t+τ) v n (t),l n (t),S(t))
where S (t) represents the status category of the time signal. J (J) n And (t+tau) is the decision-making behavior of the driver at the moment t+tau, and is the speed reduction to stop and the passing of no stop. In addition, for the convenience of calculation, v n (t)，l n (t), the value of S (t) is also discretized into several categories. In the studies herein S (t), J n (t+τ)、v n (t)、l n (t) is divided into 9, 2, 7 and 9 classes respectively. Before model training, each piece of data of each variable which is actually acquired is calculated once in each simulation step length, and the category to which the variable belongs.
Step five: a probability p that a driver passes through an intersection without stopping is calculated based on a Bayesian network model.
Firstly, extracting various variables in a Bayesian network model from collected actual data; secondly, based on category division intervals of the variables in the third step, each piece of data of each variable is corresponding to each category to form data for training; finally, a Bayesian network model is built in the Neica software, and the prepared data is imported to complete training. The results of model building and training in the software are shown in fig. 3.
Step six: under an actual simulation scene, a following model based on a Bayesian network is switched:
and (3) applying the Bayesian network decision model trained in the step five of the 4 following models under different scenes constructed in the step two to actual simulation. The flow of the following-up model switching is shown in fig. 4. When other vehicles are in front of the target vehicle and p is less than 0.5, selecting a model I as a following model of the current moment of the driver; when no other vehicle is in front of the target vehicle and p is less than 0.5, selecting the second model as a following model of the current moment of the driver; when other vehicles are in front of the target vehicle and p is more than 0.5, selecting a model III as a following model of the current moment of the driver; when no other vehicle is in front of the target vehicle and p is more than 0.5, selecting the fourth model as a following model of the current moment of the driver.
Step seven: during simulation, the step six is repeated at each moment.
Claims (5)
1. A signal intersection simulation method based on a Bayesian network and a following model is characterized by comprising the following steps of:
step one: analyzing and describing traffic scenes which a driver may face under different signal lamp states when the vehicle passes through a signalized intersection;
step two: setting up a following model of a driver in different scenes of a signalized intersection;
describing the following behaviors of the driver in different scenes by using 4 following models respectively by analyzing the behavior characteristics of the driver in different traffic scenes;
step three: establishing a decision model of a driver at a signal intersection based on a Bayesian network;
the method comprises the steps of constructing a driving behavior decision model of a signalized intersection by integrating the speed of a target vehicle, signal lamp status lamps and relevant characteristics of the distance from the target vehicle to a parking line, and taking the model as a switching basis of different following models;
step four: training a Bayesian network model based on actual intersection scene vehicle trajectory data;
step five: calculating the probability p that the driver passes through the intersection without stopping based on the Bayesian network model;
the Bayesian network dynamically calculates the probability of whether a driver can pass through an intersection (i.e. the probability of passing through a signalized intersection when decelerating and stopping or not stopping) in the current running state and the signal lamp state in real time;
step six: when other vehicles exist between the target vehicle and the parking line and p is smaller than 0.5, selecting the first model as a following model of the driver at the current moment; when no other vehicles exist between the target vehicle and the parking line and p is less than 0.5, selecting the second model as a following model of the driver at the current moment; when other vehicles exist between the target vehicle and the parking line and p is more than 0.5, selecting a model III as a following model of the driver at the current moment; when no other vehicle exists between the target vehicle and the parking line and p is more than 0.5, selecting a model IV as a following model of the driver at the current moment;
step seven: during simulation, the step six is repeated at each moment.
2. The signal fork simulation method based on the Bayesian network and the following model as claimed in claim 1, wherein:
step one the analysis and description of the vehicle passing through the signalized intersection, the driver may face 4 traffic scenes in different signal light states:
scene one: before the stop line of the intersection, the target vehicle is not the head vehicle, the current signal lamp is the red light or the green light end, and the current vehicle cannot pass through the intersection;
scene II: before the stop line of the intersection, the target vehicle is the head vehicle, the current signal lamp is the red light or the green light end, and the current vehicle cannot pass through the intersection;
scene III: before the intersection stops the line, the target vehicle is not the head vehicle, the current signal lamp is at the end of the green light or the red light, and the current vehicle can pass through the intersection without stopping;
scene four: the target vehicle is the head car before the stop line of the intersection, and the current signal lamp is the green light or the red light terminal, so that the current vehicle can pass through the intersection without stopping.
3. The method for simulating a signalized intersection based on a bayesian network and a following model according to claim 1, wherein:
step two, respectively establishing corresponding following models according to 4 different scenes which a driver at a signalized intersection possibly faces;
scene one and model one: when the target vehicle is not a head vehicle and cannot pass through the intersection without stopping, the following model of the vehicle is as follows:
Ψ(Δv n (t),l n (t))＝min{Δx n (t),l n (t)}
wherein v is n (t) is the speed of the nth vehicle at time t, Δx n (t) is the head space between the nth vehicle and the preceding vehicle (the (n-1) th vehicle), l n (t) is the distance from the nth vehicle to the stop line at time t, V op Is the optimal speed function of the vehicle under the scene, H (·) is the sea plug function, alpha 1 ，β 1 Respectively an optimal speed sensitivity coefficient and a speed difference sensitivity coefficient of a driver in the model I, wherein h is a safety distance, v max Is the maximum vehicle speed;
scene two and model two: when the target vehicle is a head vehicle and cannot pass through the intersection without stopping, the following model of the vehicle is as follows:
wherein alpha is 2 ，β 2 The optimal speed sensitivity coefficient and the speed difference sensitivity coefficient of the driver in the second model are respectively; scene three and model three: when the target vehicle is not a head vehicle and can pass through the intersection without stopping, the following model of the vehicle is as follows:
wherein alpha is 3 ，β 3 The optimal speed sensitivity coefficient and the speed difference sensitivity coefficient of the driver in the model III are respectively; scene four and model four: when the target vehicle is a head vehicle and can pass through the intersection without stopping, the following model of the vehicle is as follows:
wherein v is exp For maximum desired speed, M is any number large enough that V op (M)＝v exp 。
4. The method for simulating a signalized intersection based on a bayesian network and a following model according to claim 1, wherein:
establishing a decision model of a driver at the signalized intersection based on a Bayesian network, and establishing a Bayesian network model by taking the speed of a target vehicle, the state lamp of a signal lamp and the distance from the target vehicle to a stop line as influence factors influencing the decision of the driver based on actual traffic scene analysis, wherein the joint probability of the Bayesian network is shown in the following formula:
p[v n (t),l n (t),S(t),J n (t+τ)]
＝p(v n (t)|l n (t),S(t)).p(L n (t)).p(S(t)).p(J n (t+τ)|v n (t),l n (t),S(t))
s (t) represents the state category of the time signal lamp and is divided into 9 states in total; j (J) n (t+τ)The decision behavior of the driver at the time t+tau is represented by 0 and 1 respectively for decelerating to stop and not stopping; in addition, for the convenience of calculation, v n (t)，l n (t), the value of S (t) is also discretized into several categories.
5. The method for simulating a signalized intersection based on a bayesian network and a following model according to claim 1, wherein:
five steps of calculating the probability p that the driver does not stop passing through the intersection based on the Bayesian network model, wherein the probability p is calculated by the following formula:
wherein i, j, k represent v n (t)，l n (t), the values of S (t) respectively belong to the corresponding ith, j and k categories.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310592097.9A CN116663403B (en) | 2023-05-24 | 2023-05-24 | Signal intersection traffic simulation method based on Bayesian network and following model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310592097.9A CN116663403B (en) | 2023-05-24 | 2023-05-24 | Signal intersection traffic simulation method based on Bayesian network and following model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116663403A true CN116663403A (en) | 2023-08-29 |
| CN116663403B CN116663403B (en) | 2024-02-13 |
Family
ID=87718366
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310592097.9A Active CN116663403B (en) | 2023-05-24 | 2023-05-24 | Signal intersection traffic simulation method based on Bayesian network and following model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116663403B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170161410A1 (en) * | 2015-12-04 | 2017-06-08 | International Business Machines Corporation | System and method for simulating traffic flow distributions with approximated vehicle behavior near intersections |
| CN107103749A (en) * | 2017-05-19 | 2017-08-29 | 长安大学 | With traffic stream characteristics modeling method of speeding under car networking environment |
| CN114021982A (en) * | 2021-11-05 | 2022-02-08 | 北京航空航天大学 | Signal intersection entrance lane vehicle traffic model based on risk quantification |
| CN115620536A (en) * | 2022-10-18 | 2023-01-17 | 北京航空航天大学 | Method for improving crossing traffic efficiency based on danger degree in automatic driving environment |
-
2023
- 2023-05-24 CN CN202310592097.9A patent/CN116663403B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170161410A1 (en) * | 2015-12-04 | 2017-06-08 | International Business Machines Corporation | System and method for simulating traffic flow distributions with approximated vehicle behavior near intersections |
| CN107103749A (en) * | 2017-05-19 | 2017-08-29 | 长安大学 | With traffic stream characteristics modeling method of speeding under car networking environment |
| CN114021982A (en) * | 2021-11-05 | 2022-02-08 | 北京航空航天大学 | Signal intersection entrance lane vehicle traffic model based on risk quantification |
| CN115620536A (en) * | 2022-10-18 | 2023-01-17 | 北京航空航天大学 | Method for improving crossing traffic efficiency based on danger degree in automatic driving environment |
Non-Patent Citations (2)
| Title |
|---|
| 丁建勋;郑杨边牧;张梦婷;龙建成;: "临近交叉口的车辆跟驰换道行为研究", 交通运输系统工程与信息, no. 03 * |
| 魏允晗;韩印;: "交叉口车辆跟驰换道模型构建及仿真", 交通运输研究, no. 03 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116663403B (en) | 2024-02-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109345020B (en) | Non-signalized intersection vehicle driving behavior prediction method under complete information | |
| Murphey et al. | Driver's style classification using jerk analysis | |
| CN113050455A (en) | Digital twin test system for intelligent networked automobile and control method | |
| CN111754777A (en) | Microscopic traffic simulation method for unmanned and manned mixed traffic flow | |
| Goyani et al. | Investigation of traffic conflicts at unsignalized intersection for reckoning crash probability under mixed traffic conditions | |
| CN110751847A (en) | Decision-making method and system for automatically driving vehicle behaviors | |
| CN111813083B (en) | Scene task complexity quantitative evaluation method | |
| CN110619340B (en) | Method for generating lane change rule of automatic driving automobile | |
| CN112201070B (en) | Deep learning-based automatic driving expressway bottleneck section behavior decision method | |
| CN117012030A (en) | Urban machine non-mixed traffic signal control method based on three-dimensional macroscopic basic diagram | |
| Fujii et al. | Multi-agent based traffic simulation at merging section using coordinative behavior model | |
| Jehad et al. | Modelling and optimizing of electronic toll collection (ETC) at Malaysian toll plazas using microsimulation models | |
| CN114492043A (en) | Personalized driver following modeling method considering perception limited characteristics | |
| CN116663403B (en) | Signal intersection traffic simulation method based on Bayesian network and following model | |
| CN111476491A (en) | Vehicle motion simulation analysis method in urban traffic scene | |
| CN116588123A (en) | Risk perception early warning strategy method based on safety potential field model | |
| Van et al. | A fuzzy traffic signal control method for a single intersection under mixed traffic conditions | |
| CN112373482B (en) | Driving habit modeling method based on driving simulator | |
| CN114387778B (en) | Urban expressway congestion cause analysis method | |
| Yanagihara et al. | Latent class analysis for driving behavior on merging section | |
| Marzbanrad et al. | Prediction of driver’s accelerating behavior in the stop and go maneuvers using genetic algorithm-artificial neural network hybrid intelligence | |
| CN112396829A (en) | Intersection congestion index calculation method based on forward radar data | |
| Yavuz et al. | Performance of Autonomous Vehicles on a Bottleneck: A Simulation Study with SUMO | |
| Cheng et al. | A CA model for intrusion conflicts simulation in vehicles-bicycles laminar traffic flow | |
| Ribeiro et al. | Micro-simulation of the impact of different speeds on safety road travel and urban travel time: case study in the city of Guimarães |
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 |