Stochastic Two-Dimensional Microscopic Traffic Model: Theory and Applications

✍ Scribed by HongSheng Qi

• Publisher: Springer
• Year: 2024
• Tongue: English
• Leaves: 388
• Series: Lecture Notes in Intelligent Transportation and Infrastructure
• Category: Library

✦ Synopsis

Microscopic traffic model serves as the foundation of traffic flow theory and is the basis for applications such as traffic simulation, autonomous vehicle simulation, and digital twin technology. Conventional traffic models have primarily focused on the longitudinal dimension and have been deterministic in nature. However, vehicles' movements involve both longitudinal and lateral dimensions, and their dynamics are inherently stochastic. Therefore, a two-dimensional treatment is essential.

This book explores the theory and application of stochastic two-dimensional microscopic traffic models, including the development of theory, establishment of methods, and applications to autonomous vehicles. The book is organized into three sections: data, theory, and application. In the data section, various open-source trajectory data are analyzed and noise reduction techniques are discussed. In the theory section, various two-dimensional traffic models are developed. In the application section, the potential applications of these models are discussed, including behavioral inferences and lateral wandering. This book will be a useful reference for students, researchers and engineers in the fields of vehicle engineering and traffic technology.

✦ Table of Contents

Preface
What Does the Book Cover?
Who Is the Book For?
Which Basic Knowledge is Required to Read This Book?
Acknowledgements
Contents
Abbreviations
1 Introduction
1.1 Before We Begin
1.2 Microscopic Versus Macroscopic
1.3 The Basics of Microscopic Traffic Flow Modeling
1.4 Autonomous Vehicles and Challenge
1.5 A Review of Literature on Microscopic Traffic Flow Modeling
1.5.1 CF Models: 1D and 2D
1.5.2 Lane Changing Process Modeling
1.5.3 Lateral Dynamics
1.5.4 Data Driven Approach Toward CF and LC
1.5.5 Mixture of Theoretical Model and Data-Driven Model
1.6 Drawbacks of Current Models
1.7 Book Structure
References
2 Real World Observations, Maneuver Estimation and Behavioral Predictability
2.1 Introduction of This Chapter
2.2 Real World Observation Datasets
2.2.1 Datasets Introduction
2.2.2 Two Dimensional Noise Observation: Longitudinal
2.2.3 Two Dimensional Noise: Lateral
2.3 Maneuver Estimation of NDD: State Space Model Approach
2.3.1 Introduction
2.3.2 Description of Employed Datasets
2.3.3 Preprocessing
2.3.4 Driving Behavior State Estimation
2.4 Predicability of Driving Behavior: A Comparative Study Among Different Countries
2.4.1 Introduction
2.4.2 Permutation Entropy
2.4.3 PE Extraction from NDD
2.4.4 Case Study
2.4.5 Conclusion and Remarks
References
3 Two Dimensional Microscopic Traffic Model with Vehicle Model: Deterministic Formulation
3.1 Introduction
3.2 Intelligent Driver Model Considering Vehicle kinetic attributes
3.2.1 IDM Model
3.2.2 Model for Steering of Front Wheel
3.2.3 Two Dimensional Vehicle Dynamics Model
3.2.4 Simulation Validation
3.3 Conclusion
Reference
4 Two-Dimensional Intelligent Driving Model for Complex Road Topology
4.1 Introduction
4.2 Basic Concept of the Virtual Boundary
4.3 Derivation of Virtual Boundary Line
4.4 Two Dimensional IDM Model with Virtual Boundary Line
4.5 The Optimization of the Virtual Boundary Based on DTW
4.6 Model Analysis
4.6.1 The Location of the Virtual Boundary Line and the Convergence of the PSO
4.6.2 The Prediction of Trajectory of Maintaining at Main Stream and Exit the Highway
4.6.3 Error Analysis
References
5 A Stochastic Two-Dimensional IDM with Vehicular Dynamics
5.1 Incapability of Current Microscopic Traffic Flow Model
5.2 Objectives and Contributions
5.3 Model Development
5.3.1 Notations
5.3.2 Framework
5.3.3 Vehicle Model
5.3.4 Control Input Formulation: Acceleration
5.3.5 Control Input Formulation: Steering
5.4 Stochastic Noise
5.4.1 Stochastic Noise Type 1: OU Process
5.4.2 Synthesis of the Model
5.5 Stochastic Lyapunov Stability Analysis
5.5.1 Lyapunov Function
5.5.2 Stability Results
5.6 Numerical Results
5.6.1 Single Vehicle Acceleration from Standstill
5.6.2 A Platoon with a Leading Vehicle
5.7 Notes
References
6 Lateral Dimension Modeling and Application
6.1 Why Lateral Dimension Matters?
6.2 Background and Motivation
6.3 Literature Review
6.3.1 Lateral Movement in Lane Keeping Stage
6.3.2 Lateral Movement in Lane Changing Stage
6.4 Model Development
6.4.1 Notation List
6.4.2 Framework
6.4.3 Noise Modeling
6.4.4 Lateral Movement Modeling
6.4.5 Fokker–Planck Equation (FP Equation) of the Lateral Movement
6.4.6 LC Duration Model and the CLM Moment Distribution
6.4.7 Stability Condition
6.5 Model Calibration, and Validation
6.5.1 Calibration Method
6.5.2 Validation
6.6 Model Applications
6.6.1 Identification of the CAV
6.6.2 Lane Changing Initiation Moment (LCIE) Identification
6.6.3 The Analysis of Lane Changing Duration Distribution (LCD)
6.7 Notes
References
7 Two Dimensional Jerk Modeling: Jump-Diffusion Approach
7.1 The Definition of Jerk, Its Importance and Our Approach
7.2 Can Current Models Describ Realisitc Sped-Dependent Jerk?
7.2.1 Empirical Observation
7.2.2 Theotiretical Deduction of the Longitudinal Jerk Distribution: Markov Chain Approach
7.2.3 Results for Several Typical Car Following Models
7.2.4 Results for the Lateral Jerks
7.3 Current Practice Regarding Microscopic Modeling and Jerk Description
7.4 A Compound Poisson Approach to Car Following
7.4.1 Notations
7.4.2 Longitudinal Component
7.4.3 Lateral Component
7.4.4 Synthesis
7.5 Numerical Solution
7.5.1 Inifinitesmonial Generator
7.5.2 Numerical Scheme
7.6 Case Study
7.7 Notes
References
8 Lateral Influence on Capacity Adjustment of Lanes Number for Mixed Autonomous Vehicles
8.1 Capacity Adjustment Factor (CAF) and Current Practice
8.1.1 CAF Introduction
8.1.2 Current Practice
8.2 Symbols in the Chapter
8.3 Two Dimensional Stochastic Microscopic Traffic Model
8.3.1 Model Framework
8.3.2 HDV Two Dimensional Model
8.3.3 AV Stochastic Behavioral Model: ACC and CACC
8.3.4 Parameters of the Model
8.3.5 Example of Simulation Results
8.4 Model Validation by Replicating of the Lateral Friction Phenomenon
8.5 Experiment Design and CAF Determination Methods
8.5.1 Experiment Design
8.5.2 CAF Determination Methods
8.5.3 Results
8.6 Notes
References
9 Stochastic Lateral Wandering Patterns of Mixed Traffic Flow
9.1 Lateral Wandering and Its Implications
9.2 Real World Observations and the Work in This Chapter
9.3 Current Practive and Research Gaps
9.4 Stochastic Lateral Behavioral Model
9.4.1 Lateral Behavior Model for HDVs
9.4.2 Lateral Model for AV
9.5 Stochastic Lateral Behavioral Model Calibration
9.6 Stochastic Lateral Wandering Model
9.6.1 Pure HDV Case
9.6.2 Mixed AVs Flow Case
9.7 Stochastic Wandering Validation
9.8 Case Study
9.8.1 Scenarios Settings
9.8.2 The Influence of the MPRs
9.8.3 Dedicated Lane Case
9.9 Notes
References
10 Lateral Stochasticity in Lane Changing via Logistic Diffusion Process
10.1 Introduction
10.2 Current Practice and Drawbacks
10.3 Modeling
10.3.1 Notation and Abbreviation List
10.3.2 Deterministic Logistic Lane Changing Model
10.3.3 Logistic Diffusion Formulation of the Lane Changing Process
10.3.4 Fokker–Planck Equation (FP Equation) of the Lane Changing Trajectories
10.3.5 CLMM and LCD Distribution
10.4 Online Parameters Calibration and Inference of LCIM, CLMM and LCD
10.4.1 Parameter Calibration
10.4.2 Online LCIM Identification
10.4.3 Mean of CLMM and LCD
10.5 Case Study
10.5.1 Data Description and Preprocessing
10.5.2 LCIM Inference
10.5.3 Sensitivity of LCIM to Lane Width Argument
10.6 Notes
References
11 Game Behavior Within the Intersection
11.1 Introduction
11.2 Data Source and Processing
11.3 Model Development
11.3.1 Framework
11.3.2 SFR Model at Stopline
11.3.3 Speed Curve Model
11.3.4 Game Behavior Modeling at the Intersection Outlet
11.4 Simulation
11.4.1 Simulation Framework
11.4.2 Simulation Results
11.4.3 Application in Calculating SFR Adjustment Factor
11.5 Notes
References
12 Intersection Deadlock Probability: Stochastic Petri Net Approach
12.1 Deadlock Definition and How It Forms
12.2 Current Practice
12.3 Preliminaries on Petri Net
12.3.1 Elements of Petri Net (PN)
12.3.2 Reachability Graph
12.3.3 Symbols in the Model Development
12.4 Formulation of Intersection Traffic Dynamics as a Petri Net (PN)
12.4.1 Intersection Petri Net Framework
12.4.2 IntersectionPN Components
12.4.3 Intersection Discretization and IntersectionPN Assemble
12.5 Reachability and Reachability Graph-Based Deadlock Modeling
12.5.1 System States Definition and Deadlock Metrics
12.5.2 Reachability as an Integer Programming Problem
12.5.3 Reachability Graph Construction by Simulation
12.5.4 Markov Chain-Based Deadlock Inference
12.6 Case Study
12.6.1 Scenario Setup
12.6.2 FEFS-Based Stochastic Scenario Result Analysis
12.6.3 Pure Stochastic Scenario Result Analysis
12.6.4 Extreme Weather Scenario Result Analysis
12.7 Notes
References
13 Intersection Deadlock Protocol on Mixed Autonomous Vehicles
13.1 Deadlock and the Autonomous Vehicles
13.2 Basic Assumptions of this Chapter
13.3 Current Practice Regarding Mixed AVs Flow
13.4 Notations
13.5 Properties of Intersection Deadlock
13.5.1 Vehicle Model and the Intra-Vehicle-Blockage
13.5.2 Properties of Weak Deadlock
13.5.3 Properties of Strong Deadlock
13.6 Deadlock Detection
13.6.1 Deadlock Detection Procedure
13.6.2 HDV Model by Extended Kalman Filtering
13.6.3 Evasion Distance Propagation Algorithm-Based Weak Deadlock Detection
13.6.4 Evasion Distance Propagation Algorithm-Based Strong Deadlock Detection
13.7 Cooperative Deadlock Avoidance and Recovery Protocol
13.7.1 Deadlock Avoidance Protocol
13.7.2 Cooperative Recovery Protocol
13.8 Numeric Analysis
13.8.1 Scenario Setup
13.8.2 Illustrative Examples of the Deadlock Detection
13.8.3 Benefits of the Proposed Model
13.8.4 Leading Time for Deadlock Detection
13.8.5 The Time of Recovery Time
13.9 Notes
References
14 Closing Thoughts: Future Microscopic Modeling
14.1 Systematic Modeling of Mobility Behavior
14.2 Incorporating Advanced Technology: LLM and AIGC
14.3 Modeling Irregular Behavior: Safety and Distraction Considerations
14.4 Developing Platforms Support Simulation, Generation, Inference and More
14.5 Multi Scale Modeling